As of: 2026-10-25 23:49 CET.
Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/
Canonical page: https://postcutoff.com/models/pick/

# Which model to use

Checked: 2026-10-25. Picks last changed: 2026-10-25. (Checked daily; changed when a new model or a new leaderboard result moves a pick.)

Give this page (or its plain-text version) to an AI model instead of naming models. It can then follow instructions such as "use the frontier models from OpenAI, Google, xAI and Anthropic" or "use the best model for video understanding". Picks are curated with sources; ids, endpoints and prices come live from our registry.

How to use: map the user's words to a row below ("frontier" = the Frontier column of that lab; "best for <task>" = Pick of that task; "cheap"/"fast" = Fast or Budget; "open"/"local" = Open weights). Then call the model with the id and endpoint given. Some picks have no API (web app only, e.g. Suno): an agent that can drive a browser can still use them through the app’s website. Verify availability if the dates above are old.

Registry: https://postcutoff.com/models/index.md
Making a video? Most AI-made videos we logged were built with code (a language model writing p5.js, Three.js, Blender or Remotion), plus music and voice models, not only with video generators. Techniques, recipes and tools: https://postcutoff.com/ai-video/index.md

## By lab

### Anthropic
- Frontier: Claude Opus 5.5 (Anthropic); id `claude-opus-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $4 in / $20 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-opus-5-5/
- Fast / cheap: Claude Haiku 5.5 (Anthropic); id `claude-haiku-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $0.1 in / $0.5 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image; out: text; status current; details https://postcutoff.com/m/claude-haiku-5-5/
- Note: Opus 5.5 is #1 on the Artificial Analysis Intelligence Index (58 at max) and Anthropic's recommended default; Claude Fable 5.1 ($10/$50) is the larger 'most capable widely released' model for the hardest long-horizon work but scores 53 on AA; Claude Mythos 5.1 is the same model, invitation-only (Project Glasswing); Sonnet 5.5 (56, $2/$10) is the best value tier. Thinking is always on in all 5.5 models.

### OpenAI
- Frontier: GPT-6 Astra (OpenAI); id `gpt-6-astra` via OpenAI API, https://api.openai.com/v1/responses; $10 in / $50 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-astra/
- Fast / cheap: GPT-6 Luna (OpenAI); id `gpt-6-luna` via OpenAI API, https://api.openai.com/v1/responses; $0.1 in / $0.5 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-05; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-luna/
- Note: GPT-6 Astra ($10/$50) is the flagship (AA 53 at max); GPT-6.1 Sol ($2/$10, AA 52 at max at $0.72 per index task vs Astra $3.26) nearly matches it and is the better default for most agent work; all GPT-6 models are reasoning models with effort control. GPT-6.1 Astra was shelved. Sora 2 API is retired (2026-09-24), so OpenAI has no video API.

### Google
- Frontier: Gemini 4 Argon (Google DeepMind); id `gemini-4-argon` via Gemini API, generateContent (paid tier only; works with google_search grounding and audio input, tested 2026-10-25; the Interactions API also works), https://generativelanguage.googleapis.com/v1beta/models/gemini-4-argon:generateContent; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-4-argon/
- Fast / cheap: Gemini 3.8 Flash (Google DeepMind); id `gemini-3.8-flash` via Gemini API, https://generativelanguage.googleapis.com/v1beta/models/gemini-3.8-flash:generateContent; $0.75 in / $3.75 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-3-8-flash/
- Note: Gemini 4 Argon (GA 2026-10-13, $2/$10) leads Google's lineup and the Vals Index (68.9%), but AA still scores only its Sept 30 pre-release run (53). Gemini 3.8 Flash is the workhorse; Gemini 3.5 Flash-Lite is the cheapest and fastest (~414 tok/s on AA). Gemini 3.1 Pro is an older preview, skip it.

### xAI (SpaceXAI)
- Frontier: Grok 4.7 (xAI); id `grok-4.7` via xAI API, https://api.x.ai/v1/chat/completions; $2 in / $6 out per 1M tokens; context 500,000 tokens; knows up to 2026-05; in: text, image; out: text; status current; details https://postcutoff.com/m/grok-4-7/
- Fast / cheap: Grok 4.3 (xAI); id `grok-4.3` via xAI API, https://api.x.ai/v1/chat/completions; $1.25 in / $2.5 out per 1M tokens; context 1,000,000 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/grok-4-3/
- Note: Grok 4.7 ($2/$6, 500K context) scores 46 at xhigh on AA, well behind the US frontier; Grok 4.3 is the cheap 1M-context option; grok-build-0-1 is the coding model. xAI's media and voice models are listed by AA under 'SpaceXAI'; Grok Voice 3.0 leads AA's speech-to-speech index.

### Meta
- Frontier: Muse Wonder (Meta); id `muse-wonder-1.0` via Meta Model API, https://api.meta.ai/v1/chat/completions; $4 in / $20 out per 1M tokens; context 1,048,576 tokens; in: text, image, video, audio, pdf; out: text; status preview; details https://postcutoff.com/m/muse-wonder/
- Fast / cheap: Muse Spark 1.3 (Meta); id `muse-spark-1.3` via Meta Model API, https://api.meta.ai/v1/chat/completions; $1.25 in / $4.25 out per 1M tokens; context 1,000,000 tokens; in: text, image, video, pdf; out: text; status current; details https://postcutoff.com/m/muse-spark-1-3/
- Note: Muse Wonder (preview, $4/$20) scores 52 on AA; Muse Wonder Full scores 56 but is only in Muse Code subscriptions (no per-token API price). Muse Spark Open is the open-weights sibling (AA 39).

### DeepSeek
- Frontier: DeepSeek-V4.1-Flash (DeepSeek); id `deepseek-flash` via DeepSeek API, https://api.deepseek.com; $0.3 in / $1.2 out per 1M tokens; context 1,000,000 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/deepseek-v4-1-flash/
- Fast / cheap: DeepSeek-V4.1-Flash (DeepSeek); id `deepseek-flash` via DeepSeek API, https://api.deepseek.com; $0.3 in / $1.2 out per 1M tokens; context 1,000,000 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/deepseek-v4-1-flash/
- Note: Call as deepseek-flash (serves V4.1-Flash-1017 since Oct 17): AA 45 at max for $0.26 per index task, open weights, 1M context. DeepSeek V4 Pro (April 2026, $1.32/$3.96) is older and not scored higher on AA.

### Alibaba (Qwen)
- Frontier: Qwen3.8-Max (Alibaba (Qwen)); id `qwen3.8-max` via Alibaba Cloud Model Studio (DashScope, Singapore/Intl), https://{WorkspaceId}.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1; $2 in / $6 out per 1M tokens; context 1,000,000 tokens; in: text, image, video; out: text; status current; details https://postcutoff.com/m/qwen3-8-max/
- Fast / cheap: Qwen3.8-Flash (Alibaba (Qwen)); id `qwen3.8-flash` via Alibaba Cloud Model Studio (DashScope, Singapore/Intl), https://{WorkspaceId}.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1; $0.15 in / $0.47 out per 1M tokens; context 1,000,000 tokens; in: text, image, video; out: text; status current; details https://postcutoff.com/m/qwen3-8-flash/
- Note: Qwen3.8-Max (snapshot 0902) scores 45 on AA but costs $5.41 per index task and runs at ~40 tok/s; the registry marks it open-weights, yet AA's open-weights list did not include it (check). Alibaba's Wan 3.0 (video) and Fun-ASR (speech-to-text) are task picks below.

### Moonshot
- Frontier: Kimi K3.1 Pro (Moonshot AI); id `kimi-k3.1-pro` via Kimi API (Moonshot), https://api.moonshot.ai/v1; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/kimi-k3-1-pro/
- Fast / cheap: Kimi K2.6 (Moonshot AI); id `kimi-k2.6` via Kimi API (Moonshot), https://api.moonshot.ai/v1; $0.95 in / $4 out per 1M tokens; context 262,144 tokens; in: text, image, video; out: text; status current; details https://postcutoff.com/m/kimi-k2-6/
- Note: Kimi K3.1 Pro (2026-10-23, $2/$10) is #5 row / 4th model on AA (55), the best non-US model and the highest Terminal-Bench 4.0 AA measured (67%); weights are promised by Nov 4, 2026. Kimi K3 is the current open-weights model.

### Mistral
- Frontier: Mistral Large 4 (Mistral AI); id `mistral-large-4-0` via Mistral API, https://api.mistral.ai/v1/chat/completions; $0.68 in / $2.09 out per 1M tokens; context 1,000,000 tokens; in: text, image; out: text; status preview; details https://postcutoff.com/m/mistral-large-4/
- Fast / cheap: Mistral Small 4 (Mistral AI); id `mistral-small-2603` via Mistral API, https://api.mistral.ai/v1/chat/completions; $0.15 in / $0.6 out per 1M tokens; context 256,000 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/mistral-small-4/
- Note: Mistral Large 4 is a public preview (AA 38); weights promised by end of October 2026. Mistral's newest Voxtral 3 TTS and Voxtral Transcribe 3 models are on AA boards but not yet in our registry.

### Zhipu (Z.ai)
- Frontier: GLM-5.3 (Zhipu AI (Z.ai)); id `glm-5.3` via Z.ai API, https://api.z.ai/api/paas/v4; $1.4 in / $4.4 out per 1M tokens; context 1,000,000 tokens; in: text; out: text; status current; details https://postcutoff.com/m/glm-5-3/
- Fast / cheap: GLM-5.3-Flash / FlashX (Zhipu AI (Z.ai)); id `glm-5.3-flash` via Z.ai API, https://api.z.ai/api/paas/v4; $0.15 in / $0.5 out per 1M tokens; context 1,000,000 tokens; in: text, image, video, pdf; out: text; status current; details https://postcutoff.com/m/glm-5-3-flash/
- Note: GLM-5.3 (open weights, AA 45 at max, $2.01 per index task); GLM-5.3-Flash is open, multimodal and scores 42.

### MiniMax
- Frontier: MiniMax-M3.1-Flash (MiniMax); id `MiniMax-M3.1-Flash` via MiniMax API, https://api.minimax.io/anthropic; $0.15 in / $0.6 out per 1M tokens; context 1,000,000 tokens; in: text, image, video; out: text; status current; details https://postcutoff.com/m/minimax-m3-1-flash/
- Fast / cheap: MiniMax-M3.1-Flash (MiniMax); id `MiniMax-M3.1-Flash` via MiniMax API, https://api.minimax.io/anthropic; $0.15 in / $0.6 out per 1M tokens; context 1,000,000 tokens; in: text, image, video; out: text; status current; details https://postcutoff.com/m/minimax-m3-1-flash/
- Note: MiniMax-M3.1-Flash (2026-10-19, open weights, $0.15/$0.60) is MiniMax's top AA score (43, $0.22 per task); MiniMax H3 is also the top video model we have in the registry.

### Xiaomi
- Frontier: MiMo-V2.6-Pro (Xiaomi); id `mimo-v2.6-pro` via Xiaomi MiMo API; $0.435 in / $0.87 out per 1M tokens; context 1,000,000 tokens; in: text, image, video, audio; out: text; status current; details https://postcutoff.com/m/mimo-v2-6-pro/
- Fast / cheap: MiMo-V2.6-Flash (Xiaomi); id `mimo-v2.6-flash` via Xiaomi MiMo API; $0.14 in / $0.28 out per 1M tokens; context 1,000,000 tokens; in: text, image, video, audio; out: text; status current; details https://postcutoff.com/m/mimo-v2-6-flash/
- Note: MiMo-V2.6-Pro is the #1 open-weights model on AA (46), MIT licence, omnimodal input (text, image, video, audio), $0.435/$0.87.

## By task

### General reasoning and knowledge (best overall model) [general]
- Pick: Claude Opus 5.5 (Anthropic); id `claude-opus-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $4 in / $20 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-opus-5-5/
- Alternative: Kimi K3.1 Pro (Moonshot AI); id `kimi-k3.1-pro` via Kimi API (Moonshot), https://api.moonshot.ai/v1; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/kimi-k3-1-pro/
- Alternative: GPT-6 Astra (OpenAI); id `gpt-6-astra` via OpenAI API, https://api.openai.com/v1/responses; $10 in / $50 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-astra/
- Alternative: Gemini 4 Argon (Google DeepMind); id `gemini-4-argon` via Gemini API, generateContent (paid tier only; works with google_search grounding and audio input, tested 2026-10-25; the Interactions API also works), https://generativelanguage.googleapis.com/v1beta/models/gemini-4-argon:generateContent; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-4-argon/
- Alternative: Claude Sonnet 5.5 (Anthropic); id `claude-sonnet-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $2 in / $10 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-sonnet-5-5/
- Open weights: MiMo-V2.6-Pro (Xiaomi); id `mimo-v2.6-pro` via Xiaomi MiMo API; $0.435 in / $0.87 out per 1M tokens; context 1,000,000 tokens; in: text, image, video, audio; out: text; status current; details https://postcutoff.com/m/mimo-v2-6-pro/
- Budget: GPT-6.1 Sol (OpenAI); id `gpt-6.1-sol` via OpenAI API, https://api.openai.com/v1/responses; $2 in / $10 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-1-sol/
- Why: Artificial Analysis Intelligence Index: Claude Opus 5.5 58 (max), Sonnet 5.5 56, Muse Wonder Full 56 (no API price), Kimi K3.1 Pro 55, Fable 5.1 / GPT-6 Astra / Gemini 4 Argon 53 (Argon measured on its Sept 30 pre-release). GPT-6.1 Sol reaches 52 at $0.72 per index task vs $5.98 for Opus 5.5 at max. Vals Index disagrees on the order: Argon 68.9% first, Sonnet 5.5 67.0%, Opus 5.5 67.0%.
- Sources: https://artificialanalysis.ai/leaderboards/models , https://deepmind.google/models/model-cards/gemini-4-argon/ , https://postcutoff.com/e/2026-10-23-kimi-k3-1-pro/ , https://postcutoff.com/e/2026-10-12-meta-muse-wonder/

### Coding and agentic software work [coding]
- Pick: Claude Opus 5.5 (Anthropic); id `claude-opus-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $4 in / $20 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-opus-5-5/
- Alternative: Claude Sonnet 5.5 (Anthropic); id `claude-sonnet-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $2 in / $10 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-sonnet-5-5/
- Alternative: GPT-6 Astra (OpenAI); id `gpt-6-astra` via OpenAI API, https://api.openai.com/v1/responses; $10 in / $50 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-astra/
- Alternative: Kimi K3.1 Pro (Moonshot AI); id `kimi-k3.1-pro` via Kimi API (Moonshot), https://api.moonshot.ai/v1; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/kimi-k3-1-pro/
- Alternative: Gemini 4 Argon (Google DeepMind); id `gemini-4-argon` via Gemini API, generateContent (paid tier only; works with google_search grounding and audio input, tested 2026-10-25; the Interactions API also works), https://generativelanguage.googleapis.com/v1beta/models/gemini-4-argon:generateContent; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-4-argon/
- Open weights: MiMo-V2.6-Pro (Xiaomi); id `mimo-v2.6-pro` via Xiaomi MiMo API; $0.435 in / $0.87 out per 1M tokens; context 1,000,000 tokens; in: text, image, video, audio; out: text; status current; details https://postcutoff.com/m/mimo-v2-6-pro/
- Budget: Claude Sonnet 5.5 (Anthropic); id `claude-sonnet-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $2 in / $10 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-sonnet-5-5/
- Why: Opus 5.5 (vendor): SWE-bench Pro 89.9%, Terminal-Bench 4.0 66.4% (Astra 57.9-58.2%, Argon 57.4%), FrontierCode v1.1 54.4%. Sources disagree on Anthropic's order: Sonnet 5.5 reports 70.6% on Terminal-Bench 4.0 and scores 68 on AA's Coding Agents Index in Claude Code vs 66 for Opus 5.5; Kimi K3.1 Pro has the highest Terminal-Bench 4.0 AA measured (67%); Argon leads DeepSWE v1.1 (77.9%) and Astra leads FrontierSWE v2 (65.5%). Open: MiMo-V2.6-Pro, vendor DeepSWE v1.1 71.9.
- Sources: https://postcutoff.com/e/2026-09-22-claude-opus-5-5/ , https://postcutoff.com/e/2026-09-28-claude-sonnet-5-5/ , https://deepmind.google/models/model-cards/gemini-4-argon/ , https://postcutoff.com/e/2026-10-12-meta-muse-wonder/ , https://postcutoff.com/m/mimo-v2-6-pro/

### Long documents and long context (500K-1M tokens) [long-context]
- Pick: Gemini 4 Argon (Google DeepMind); id `gemini-4-argon` via Gemini API, generateContent (paid tier only; works with google_search grounding and audio input, tested 2026-10-25; the Interactions API also works), https://generativelanguage.googleapis.com/v1beta/models/gemini-4-argon:generateContent; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-4-argon/
- Alternative: GPT-6 Astra (OpenAI); id `gpt-6-astra` via OpenAI API, https://api.openai.com/v1/responses; $10 in / $50 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-astra/
- Alternative: Claude Opus 5.5 (Anthropic); id `claude-opus-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $4 in / $20 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-opus-5-5/
- Alternative: Kimi K3.1 Pro (Moonshot AI); id `kimi-k3.1-pro` via Kimi API (Moonshot), https://api.moonshot.ai/v1; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/kimi-k3-1-pro/
- Open weights: DeepSeek-V4.1-Flash (DeepSeek); id `deepseek-flash` via DeepSeek API, https://api.deepseek.com; $0.3 in / $1.2 out per 1M tokens; context 1,000,000 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/deepseek-v4-1-flash/
- Budget: Grok 4.3 (xAI); id `grok-4.3` via xAI API, https://api.x.ai/v1/chat/completions; $1.25 in / $2.5 out per 1M tokens; context 1,000,000 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/grok-4-3/
- Why: Google's model card: GraphWalks BFS 256K-1M F1 84.2 for Argon vs 71.8 GPT-6 Astra, 66.8 Claude Opus 5.5, 65.0 Fable 5.1 (vendor table); Argon also decodes up to ~1M output tokens via long decode. Budget: Grok 4.3 offers 1M context at $1.25/$2.50; open: DeepSeek V4.1 Flash, 1M context, AA 45.
- Sources: https://deepmind.google/models/model-cards/gemini-4-argon/ , https://postcutoff.com/m/gemini-4-argon/ , https://postcutoff.com/m/grok-4-3/

### Mathematics (competition and research-level, proofs) [math]
- Pick: GPT-6 Astra (OpenAI); id `gpt-6-astra` via OpenAI API, https://api.openai.com/v1/responses; $10 in / $50 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-astra/
- Alternative: Gemini 4 Argon (Google DeepMind); id `gemini-4-argon` via Gemini API, generateContent (paid tier only; works with google_search grounding and audio input, tested 2026-10-25; the Interactions API also works), https://generativelanguage.googleapis.com/v1beta/models/gemini-4-argon:generateContent; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-4-argon/
- Alternative: Claude Opus 5.5 (Anthropic); id `claude-opus-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $4 in / $20 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-opus-5-5/
- Alternative: Claude Fable 5.1 (Anthropic); id `claude-fable-5-1` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $10 in / $50 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-fable-5-1/
- Open weights: DeepSeek-V4.1-Flash (DeepSeek); id `deepseek-flash` via DeepSeek API, https://api.deepseek.com; $0.3 in / $1.2 out per 1M tokens; context 1,000,000 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/deepseek-v4-1-flash/
- Budget: GPT-6.1 Sol (OpenAI); id `gpt-6.1-sol` via OpenAI API, https://api.openai.com/v1/responses; $2 in / $10 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-1-sol/
- Why: GPT-6 Astra: FrontierMath Tier 4 97.6% (OpenAI tables) and a string of Lean-verified open-problem results in Epoch's FrontierMath Erdős runs (dataset entries). Google's card has Argon ahead on RiemannBench (76.0 vs Astra 72.0, Opus 5.5 69.6). For Lean formalization use leanstral-1-5 (Mistral, open). Open pick is by overall AA index; MathArena Apex 65.6 is vendor-reported for DeepSeek V4.1 Flash.
- Sources: https://postcutoff.com/e/2026-09-03-gpt-6-astra/ , https://postcutoff.com/e/2026-09-03-frontiermath-erdos-astra-disproves-erdos-problem-1/ , https://deepmind.google/models/model-cards/gemini-4-argon/ , https://postcutoff.com/e/2026-09-10-deepseek-v4-1-flash/

### Image understanding (charts, documents, screenshots, photos) [vision]
- Pick: Gemini 4 Argon (Google DeepMind); id `gemini-4-argon` via Gemini API, generateContent (paid tier only; works with google_search grounding and audio input, tested 2026-10-25; the Interactions API also works), https://generativelanguage.googleapis.com/v1beta/models/gemini-4-argon:generateContent; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-4-argon/
- Alternative: GPT-6 Astra (OpenAI); id `gpt-6-astra` via OpenAI API, https://api.openai.com/v1/responses; $10 in / $50 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-astra/
- Alternative: Claude Opus 5.5 (Anthropic); id `claude-opus-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $4 in / $20 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-opus-5-5/
- Open weights: MiMo-V2.6-Pro (Xiaomi); id `mimo-v2.6-pro` via Xiaomi MiMo API; $0.435 in / $0.87 out per 1M tokens; context 1,000,000 tokens; in: text, image, video, audio; out: text; status current; details https://postcutoff.com/m/mimo-v2-6-pro/
- Budget: Gemini 3.8 Flash (Google DeepMind); id `gemini-3.8-flash` via Gemini API, https://generativelanguage.googleapis.com/v1beta/models/gemini-3.8-flash:generateContent; $0.75 in / $3.75 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-3-8-flash/
- Why: Only vendor numbers checked: Chartography 71.6 Argon vs 71.0 Astra, 66.3 Opus 5.5, 46.2 Fable 5.1 (Google's table). We could not read Arena's Vision leaderboard (403) or a current MMMU board. For OCR of PDFs and scans, mistral-ocr-4-1 is a dedicated option. Open pick by overall index with native vision; not benchmarked on vision here.
- Sources: https://deepmind.google/models/model-cards/gemini-4-argon/ , https://artificialanalysis.ai/models/open-source

### Video understanding (watching video natively, long video) [video-understanding]
- Pick: Gemini 4 Argon (Google DeepMind); id `gemini-4-argon` via Gemini API, generateContent (paid tier only; works with google_search grounding and audio input, tested 2026-10-25; the Interactions API also works), https://generativelanguage.googleapis.com/v1beta/models/gemini-4-argon:generateContent; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-4-argon/
- Alternative: Gemini 3.8 Flash (Google DeepMind); id `gemini-3.8-flash` via Gemini API, https://generativelanguage.googleapis.com/v1beta/models/gemini-3.8-flash:generateContent; $0.75 in / $3.75 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-3-8-flash/
- Alternative: Qwen3.8-Omni-Flash (Alibaba (Qwen)); id `qwen3.8-omni-flash` via Alibaba Cloud Model Studio (DashScope, Singapore/Intl), https://{WorkspaceId}.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1; $0.15 in / $0.47 out per 1M tokens; context 1,000,000 tokens; in: text, image, audio, video; out: text; status current; details https://postcutoff.com/m/qwen3-8-omni-flash/
- Alternative: Kimi K3 (Moonshot AI); id `kimi-k3` via Kimi API (Moonshot), https://api.moonshot.ai/v1; $3 in / $15 out per 1M tokens; context 1,048,576 tokens; in: text, image, video; out: text; status current; details https://postcutoff.com/m/kimi-k3/
- Open weights: MiMo-V2.6-Pro (Xiaomi); id `mimo-v2.6-pro` via Xiaomi MiMo API; $0.435 in / $0.87 out per 1M tokens; context 1,000,000 tokens; in: text, image, video, audio; out: text; status current; details https://postcutoff.com/m/mimo-v2-6-pro/
- Budget: Gemini 3.5 Flash-Lite (Google DeepMind); id `gemini-3.5-flash-lite` via Gemini API, https://generativelanguage.googleapis.com/v1beta/models/gemini-3.5-flash-lite:generateContent; $0.3 in / $2.5 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-3-5-flash-lite/
- Why: LVBench (Google's table): Argon 91.7 vs GPT-6 Astra 87.5, Opus 5.5 83.7, Fable 5.1 79.7; Gemini 3.8 Flash 87.8% on long-video understanding in agentic mode (model card). OpenAI and Anthropic APIs take images, not video, so Gemini is the native option; Qwen3.8-Omni-Flash and MiMo-V2.6 take video plus audio. No independent Video-MME board for these models was found.
- Sources: https://deepmind.google/models/model-cards/gemini-4-argon/ , https://postcutoff.com/m/gemini-3-8-flash/ , https://postcutoff.com/m/qwen3-8-omni-flash/

### Image generation (text to image) [image-generation]
- Pick: GPT Image 2.5 Sunburst (OpenAI); id `gpt-image-2.5-sunburst` via OpenAI API, https://api.openai.com/v1/images/generations; text input $5, text cached input $1.25, image input $8, image cached input $2, image output $30 (per 1M tokens (USD)); in: text, image; out: image; status current; details https://postcutoff.com/m/gpt-image-2-5-sunburst/
- Alternative: GPT Image 2.5 Flare (OpenAI); id `gpt-image-2.5-flare` via OpenAI API, https://api.openai.com/v1/images/generations; text input $5, text cached input $1.25, image input $8, image cached input $2, image output $30 (per 1M tokens (USD)); in: text, image; out: image; status current; details https://postcutoff.com/m/gpt-image-2-5-flare/
- Alternative: Nano Banana 2.1 (Gemini Nano Banana 2.1) (Google DeepMind); id `gemini-nano-banana-2.1` via Gemini API, https://generativelanguage.googleapis.com/v1beta/models/gemini-nano-banana-2.1:generateContent; $1.5 in / $7.5 out per 1M tokens; context 131,072 tokens; knows up to 2026-03; in: text, image, video, pdf; out: image, text; status current; details https://postcutoff.com/m/gemini-nano-banana-2-1/
- Alternative: MAI-Image-2.6 (Microsoft); id `MAI-Image-2.6` via Microsoft Foundry (Foundry Models sold by Azure), https://<resource-name>.services.ai.azure.com/mai/v1/images/generations; text input $5, image input $8, image output $38 (USD per 1M tokens (OpenRouter listing, provider Azure); Artificial Analysis estimates ~$38.9 per 1,000 images at 1024x1024); in: text, image; out: image; status preview; details https://postcutoff.com/m/mai-image-2-6/
- Alternative: Grok Imagine Image 2.0 (xAI); id `grok-imagine-image-2.0` via xAI API, https://api.x.ai/v1/images/generations; per image $0.04 (per output image (base tier per xAI models page)); in: text, image; out: image; status current; details https://postcutoff.com/m/grok-imagine-image-2/
- Alternative: Muse Image (Meta); id `muse-image-1.0` via Meta Model API, https://api.meta.ai/v1/images/generations; per image $0.01 (USD per generated image (Meta Model API; search grounding included)); in: text, image; out: image; status current; details https://postcutoff.com/m/muse-image/
- Open weights: FLUX 3 Image Klein (Black Forest Labs); id `flux-3-image-klein` via BFL API; per image $0.009 (per 1K image ($0.008 at 768², $0.011 at 1.5K, $0.014 at 2K, $0.036 at 4K; generation and editing priced the same)); in: text, image; out: image; status current; details https://postcutoff.com/m/flux-3-image-klein/
- Budget: Muse Image (Meta); id `muse-image-1.0` via Meta Model API, https://api.meta.ai/v1/images/generations; per image $0.01 (USD per generated image (Meta Model API; search grounding included)); in: text, image; out: image; status current; details https://postcutoff.com/m/muse-image/
- Why: Artificial Analysis Text-to-Image (AA-Image-T2I v2.1): GPT Image 2.5 Sunburst 1222, Flare 1207, GPT Image 2 1186, Nano Banana 2.1 1176 at $33.6 per 1k images vs $210.7 for Sunburst, MAI-Image-2.6 1166, Grok Imagine Image 2.0 1166. AA does not mark open weights; FLUX 3 Image Klein (5.6B, native 4K) and qwen-image-2-1 are the newest open options, unranked here. For typography, ideogram-4.
- Sources: https://artificialanalysis.ai/text-to-image/arena/leaderboard-text

### Image editing (instruction edits, inpainting, multi-reference) [image-editing]
- Pick: GPT Image 2.5 Sunburst (OpenAI); id `gpt-image-2.5-sunburst` via OpenAI API, https://api.openai.com/v1/images/generations; text input $5, text cached input $1.25, image input $8, image cached input $2, image output $30 (per 1M tokens (USD)); in: text, image; out: image; status current; details https://postcutoff.com/m/gpt-image-2-5-sunburst/
- Alternative: GPT Image 2.5 Flare (OpenAI); id `gpt-image-2.5-flare` via OpenAI API, https://api.openai.com/v1/images/generations; text input $5, text cached input $1.25, image input $8, image cached input $2, image output $30 (per 1M tokens (USD)); in: text, image; out: image; status current; details https://postcutoff.com/m/gpt-image-2-5-flare/
- Alternative: Nano Banana 2.1 (Gemini Nano Banana 2.1) (Google DeepMind); id `gemini-nano-banana-2.1` via Gemini API, https://generativelanguage.googleapis.com/v1beta/models/gemini-nano-banana-2.1:generateContent; $1.5 in / $7.5 out per 1M tokens; context 131,072 tokens; knows up to 2026-03; in: text, image, video, pdf; out: image, text; status current; details https://postcutoff.com/m/gemini-nano-banana-2-1/
- Open weights: Qwen-Image-2.1 (Alibaba (Qwen)); id `Qwen/Qwen-Image-2.1` via Hugging Face; price not recorded (check the provider); in: text, image; out: image; status current; details https://postcutoff.com/m/qwen-image-2-1/
- Budget: Nano Banana 2.1 (Gemini Nano Banana 2.1) (Google DeepMind); id `gemini-nano-banana-2.1` via Gemini API, https://generativelanguage.googleapis.com/v1beta/models/gemini-nano-banana-2.1:generateContent; $1.5 in / $7.5 out per 1M tokens; context 131,072 tokens; knows up to 2026-03; in: text, image, video, pdf; out: image, text; status current; details https://postcutoff.com/m/gemini-nano-banana-2-1/
- Why: Artificial Analysis Image Editing: GPT Image 2.5 Sunburst 1199, Flare 1182, MAI-Image-2.6 1151, Krea 3 Ultra 1151, Nano Banana 2.1 1151 ($33.6 vs $210.7 per 1k images). Qwen-Image-2.1 (open, research licence) does multi-reference and RGBA edits; not on AA's top 10.
- Sources: https://artificialanalysis.ai/image/leaderboard/editing , https://postcutoff.com/m/qwen-image-2-1/

### Video generation (text to video, with audio) [video-generation]
- Pick: Wan 3.0 (Alibaba Cloud (Tongyi Lab)); id `wan3.0-video` via Alibaba Cloud Model Studio (DashScope), https://{WorkspaceId}.ap-southeast-1.maas.aliyuncs.com/api/v1/services/aigc/video-generation/video-synthesis; per second 480p $0.05, per second 720p $0.1, per second 1080p $0.2 (USD per second, Singapore (International) region list price for wan3.0-video, with a limited-time 30% discount shown; other regions 0.041/0.083/0.165 per s. Prime (faster) variant wan3.0-video-prime 0.068/0.14/0.28 per s in Singapore. Input + output video duration both billed; failed requests free); in: text, image, video, audio, pdf; out: video, audio; status current; details https://postcutoff.com/m/wan-3-0/
- Alternative: HeyGen Video 1 (HeyGen); id `heygen-video-1` via HeyGen API, https://api.heygen.com/v3/models/videos; t2v i2v per second 480p $0.01, t2v i2v per second 768p $0.018, t2v i2v per second 1080p 2k $0.05, ref2v per second 480p $0.035, ref2v per second 768p $0.05, ref2v per second 1080p 2k $0.09 (USD per second of output video, API launch promotion (50% off; HeyGen says list price is double these rates)); in: text, image, video, audio; out: video, audio; status current; details https://postcutoff.com/m/heygen-video-1/
- Alternative: Dreamina Seedance 2.5 (ByteDance Seed); id `dreamina-seedance-2-5-260628` via BytePlus ModelArk, https://ark.ap-southeast.bytepluses.com/api/v3/; per 1m tokens 480p 720p $10.7, per 1m tokens 1080p $11.7, per 1m tokens with video input 480p 720p $6.4, per 1m tokens with video input 1080p $7 (USD per 1M video tokens on BytePlus ModelArk (online inference; 4K not supported). Tokens = (input + output video seconds) x width x height x fps / 1024. BytePlus examples, 5 s 16:9 without video input: 480p 0.514 (0.103/s), 720p 1.156 (0.231/s), 1080p 2.843 (0.569/s). Draft mode is billed as 480p); in: text, image, video, audio; out: video, audio; status current; details https://postcutoff.com/m/seedance-2-5/
- Alternative: FLUX 3 Max (Black Forest Labs); id `blackforestlabs/flux-3-max/text-to-video` via fal; per second 720p $0.0425, per second 1080p $0.0725 (USD per second of output video on fal, 75% launch discount (Oct 15 - Nov 7, 2026); list price after Nov 7: 0.17/s 720p, 0.29/s 1080p. Text-to-video and image-to-video same rate); in: text, image; out: video, audio; status current; details https://postcutoff.com/m/flux-3-max/
- Alternative: MiniMax H3 (MiniMax); id `MiniMax-H3` via MiniMax API (Video Generation V2), https://api.minimax.io; per second 768p $0.08, per second 2k $0.13 (per second of output video (USD). H3-Max (fal.ai post-trained, fast): 480P 0.05/s, 768P 0.08/s. Extra input images 0.04 each after 5 free); in: text, image, video, audio; out: video, audio; status current; details https://postcutoff.com/m/minimax-h3/
- Alternative: Gemini Omni Flash (Omni 1.1 Flash) (Google DeepMind); id `gemini-omni-1.1-flash` via Gemini API (Interactions API), https://generativelanguage.googleapis.com/v1beta/interactions; $1.5 in / $9 out per 1M tokens; context 1,048,576 tokens; in: text, image, audio, video; out: video, audio; status current; details https://postcutoff.com/m/gemini-omni-flash/
- Alternative: Utopai X (Utopai Studios); no API id recorded; credits per second 1080p $93 (PAI credits per second of 1080p video (no API). Plans 15-379 USD/month for 1,000-35,000 credits; top-up 15 USD per 1,000 credits (so about 1.40 USD/s at the top-up rate, our arithmetic)); in: text, image; out: video, audio; status current; details https://postcutoff.com/m/utopai-x/
- Open weights: MiniMax H3 (MiniMax); id `MiniMax-H3` via MiniMax API (Video Generation V2), https://api.minimax.io; per second 768p $0.08, per second 2k $0.13 (per second of output video (USD). H3-Max (fal.ai post-trained, fast): 480P 0.05/s, 768P 0.08/s. Extra input images 0.04 each after 5 free); in: text, image, video, audio; out: video, audio; status current; details https://postcutoff.com/m/minimax-h3/
- Budget: Dreamina Seedance 2.0 (ByteDance Seed); id `dreamina-seedance-2-0-260128` via BytePlus ModelArk, https://ark.ap-southeast.bytepluses.com/api/v3/; per 1m tokens 480p 720p $7, per 1m tokens 1080p $7.7, per 1m tokens 4k $4, per 1m tokens with video input 480p 720p $4.3, per 1m tokens with video input 1080p $4.7 (USD per 1M video tokens on BytePlus ModelArk (online inference). BytePlus examples, 5 s 16:9 without video input: 480p 0.35 (0.07/s), 720p 0.76 (0.15/s), 1080p 1.87 (0.37/s), 4K 3.89 (0.78/s). Fast variant 5.6/1M (480p/720p only), Mini 3.5/1M); in: text, image, video, audio; out: video, audio; status current; details https://postcutoff.com/m/seedance-2-0/
- Why: Artificial Analysis text-to-video arena (AA-Video-T2V v2.1, read 2026-10-25): Wan 3.0 #1 (Elo 1175, $12/min), Utopai X #2 (1167, web app only, no API), HeyGen Video 1 #3 (1160), MiniMax H3 #4 (1158, open weights), Seedance 2.5 #5 (1158), FLUX 3 Max #6 (1155), Seedance 2.0 #9 (1136), Gemini Omni Flash 1.1 #10 (1130, $9.12/min), Kling 3.0 Omni Pro #17 (1020), Kling 3.0 Pro #18 (1000, the board's anchor), Veo 3.1 #21 (963, $24/min). Kling 4.0 (Oct 15; up to 30 s, 15 references, 10 keyframes) is not on the board yet; it is on fal (kling-4-0). Wan 3.0 has an API (Alibaba Model Studio, OpenRouter). Starting from an image or a character: see image-to-video. Multi-model hubs let one key reach many of these: fal (per-second billing, ids in our registry files) and Higgsfield (web app with Kling, Veo 3.1, Seedance 2.5, Gemini Omni Flash and others; official CLI: npm install -g @higgsfield/cli, v1.2.2 on npm, repo higgsfield-ai/cli). Seedance 2.5 is the model named most often by creators in our video log (30 videos), usually with a language model writing the shot prompts. For many videos, code-drawn visuals beat generated footage: see /ai-video/.
- Sources: https://artificialanalysis.ai/video/leaderboard/text-to-video , https://postcutoff.com/m/wan-3-0/ , https://postcutoff.com/m/kling-4-0/ , https://www.npmjs.com/package/@higgsfield/cli , https://github.com/higgsfield-ai/cli , https://postcutoff.com/ai-video/

### Image to video (animate a start frame, keep the character or product consistent) [image-to-video]
- Pick: MiniMax H3 (MiniMax); id `MiniMax-H3` via MiniMax API (Video Generation V2), https://api.minimax.io; per second 768p $0.08, per second 2k $0.13 (per second of output video (USD). H3-Max (fal.ai post-trained, fast): 480P 0.05/s, 768P 0.08/s. Extra input images 0.04 each after 5 free); in: text, image, video, audio; out: video, audio; status current; details https://postcutoff.com/m/minimax-h3/
- Alternative: Dreamina Seedance 2.5 (ByteDance Seed); id `dreamina-seedance-2-5-260628` via BytePlus ModelArk, https://ark.ap-southeast.bytepluses.com/api/v3/; per 1m tokens 480p 720p $10.7, per 1m tokens 1080p $11.7, per 1m tokens with video input 480p 720p $6.4, per 1m tokens with video input 1080p $7 (USD per 1M video tokens on BytePlus ModelArk (online inference; 4K not supported). Tokens = (input + output video seconds) x width x height x fps / 1024. BytePlus examples, 5 s 16:9 without video input: 480p 0.514 (0.103/s), 720p 1.156 (0.231/s), 1080p 2.843 (0.569/s). Draft mode is billed as 480p); in: text, image, video, audio; out: video, audio; status current; details https://postcutoff.com/m/seedance-2-5/
- Alternative: FLUX 3 Max (Black Forest Labs); id `blackforestlabs/flux-3-max/text-to-video` via fal; per second 720p $0.0425, per second 1080p $0.0725 (USD per second of output video on fal, 75% launch discount (Oct 15 - Nov 7, 2026); list price after Nov 7: 0.17/s 720p, 0.29/s 1080p. Text-to-video and image-to-video same rate); in: text, image; out: video, audio; status current; details https://postcutoff.com/m/flux-3-max/
- Alternative: Gemini Omni Flash (Omni 1.1 Flash) (Google DeepMind); id `gemini-omni-1.1-flash` via Gemini API (Interactions API), https://generativelanguage.googleapis.com/v1beta/interactions; $1.5 in / $9 out per 1M tokens; context 1,048,576 tokens; in: text, image, audio, video; out: video, audio; status current; details https://postcutoff.com/m/gemini-omni-flash/
- Alternative: Wan 3.0 (Alibaba Cloud (Tongyi Lab)); id `wan3.0-video` via Alibaba Cloud Model Studio (DashScope), https://{WorkspaceId}.ap-southeast-1.maas.aliyuncs.com/api/v1/services/aigc/video-generation/video-synthesis; per second 480p $0.05, per second 720p $0.1, per second 1080p $0.2 (USD per second, Singapore (International) region list price for wan3.0-video, with a limited-time 30% discount shown; other regions 0.041/0.083/0.165 per s. Prime (faster) variant wan3.0-video-prime 0.068/0.14/0.28 per s in Singapore. Input + output video duration both billed; failed requests free); in: text, image, video, audio, pdf; out: video, audio; status current; details https://postcutoff.com/m/wan-3-0/
- Alternative: Kling 4.0 (and 4.0 Flash) (Kuaishou); id `kling/v4/image-to-video` via fal.ai, https://queue.fal.run/kling/v4/image-to-video; per second 720p $0.157, per second 1080p $0.235 (USD per second of output on fal (kling/v4/text-to-video, image-to-video and reference-to-video, read 2026-10-25); reference-to-video with video references $0.094 / $0.141 per s, billed on input plus output length (minimum 2x the output). Kling's own API price not found); in: text, image, video; out: video, audio; status current; details https://postcutoff.com/m/kling-4-0/
- Alternative: Kling 3.0 (VIDEO 3.0 / 3.0 Omni) (Kuaishou); id `fal-ai/kling-video/v3/standard/text-to-video` via fal.ai; standard per second $0.084, standard per second audio $0.126, pro per second $0.112, pro per second audio $0.168 (USD per second of output on fal (fal-ai/kling-video/v3/standard and /v3/pro, audio off / audio on, read 2026-10-25). Artificial Analysis lists Kling 3.0 1080p (Pro) at $10.08 per minute and 3.0 Omni 1080p (Pro) at $8.40 per minute on Kling's own API); in: text, image, video; out: video, audio; status current; details https://postcutoff.com/m/kling-3-0/
- Open weights: MiniMax H3 (MiniMax); id `MiniMax-H3` via MiniMax API (Video Generation V2), https://api.minimax.io; per second 768p $0.08, per second 2k $0.13 (per second of output video (USD). H3-Max (fal.ai post-trained, fast): 480P 0.05/s, 768P 0.08/s. Extra input images 0.04 each after 5 free); in: text, image, video, audio; out: video, audio; status current; details https://postcutoff.com/m/minimax-h3/
- Budget: FLUX 3 Max (Black Forest Labs); id `blackforestlabs/flux-3-max/text-to-video` via fal; per second 720p $0.0425, per second 1080p $0.0725 (USD per second of output video on fal, 75% launch discount (Oct 15 - Nov 7, 2026); list price after Nov 7: 0.17/s 720p, 0.29/s 1080p. Text-to-video and image-to-video same rate); in: text, image; out: video, audio; status current; details https://postcutoff.com/m/flux-3-max/
- Why: Artificial Analysis image-to-video arena (AA-Video-I2V v1.1, with audio, read 2026-10-25): MiniMax H3 768p #1 (Elo 1140, $4.80/min, open weights), FLUX 3 Max 720p #2 (1138, $2.55/min at fal's launch discount to Nov 7), Seedance 2.5 #3 (1138, $34.12/min), Gemini Omni Flash 1.1 #4 (1132, $9.12/min), Wan 3.0 #5 (1127), MiniMax H3 Max #6 (1125), Seedance 2.0 #7 (1121); AA's rank ranges put the top 7 within error of one another. Then Veo 3.1 #13 (1028), Kling 3.0 Omni Pro #17 (1004), Kling 3.0 Pro #18 (1000, anchor). The arena measures overall preference, not identity consistency specifically, and Kling 4.0 is not ranked yet. One user's single-shot test (anecdotal, not the basis of this pick): Kling 4.0 and Kling 3.0 Pro held the character best, Seedance 2.5 had the liveliest motion, Gemini Omni Flash changed the character most. For a recurring character, pass the same reference image (or first and last frame) every time; Kling 4.0 (up to 15) and Wan 3.0 (up to 20) accept several references per request.
- Sources: https://artificialanalysis.ai/video/leaderboard/image-to-video , https://postcutoff.com/m/minimax-h3/ , https://postcutoff.com/m/kling-4-0/ , https://fal.ai/models/kling/v4/image-to-video

### Text-to-speech (voice generation, narration) [text-to-speech]
- Pick: Eleven v4 / Eleven v4 Turbo (ElevenLabs); id `eleven_v4` via ElevenLabs API, https://api.elevenlabs.io/v1/text-to-speech/{voice_id}; per 1k characters $0.08, turbo per 1k characters $0.04 (per 1K characters, list price for eleven_v4 (eleven_v4_turbo $0.04/1K). The 72% launch promo ($0.022 / $0.011 per 1K) ended on 2026-10-12; the pricing page showed list prices with no discount label when checked on 2026-10-14); in: text; out: audio; status current; details https://postcutoff.com/m/elevenlabs-v4/
- Alternative: Qwen-Audio-3.1-TTS (Plus / Flash) (Alibaba (Qwen)); id `qwen-audio-3.1-tts-plus` via Alibaba Cloud Model Studio / QwenCloud (Singapore, Beijing), https://{WorkspaceId}.ap-southeast-1.maas.aliyuncs.com/api/v1/services/audio/tts/SpeechSynthesizer; per 10k characters intl $0.3, per 10k characters beijing $0.2 (USD per 10,000 characters for qwen-audio-3.1-tts-plus, no output charge ($30 / $20 per 1M characters, Singapore / Beijing). qwen-audio-3.1-tts-flash is billed by tokens: $0.23 in / $1.87 out per 1M tokens (Singapore), $0.212 / $1.696 (Beijing)); in: text, audio; out: audio; status current; details https://postcutoff.com/m/qwen-audio-3-1-tts-plus/
- Alternative: Cartesia Sonic-3.6 (Cartesia); id `sonic-3.6` via Cartesia API, https://api.cartesia.ai/tts/bytes (also /tts/sse and WebSocket); per 1k characters $0.037 (approx., derived: ~1 credit per character; Scale plan $299/mo for 8M credits (Startup $49 for 1.25M ≈ $0.039/1K). Artificial Analysis lists $49 per 1M chars); in: text; out: audio; status current; details https://postcutoff.com/m/cartesia-sonic-3-6/
- Alternative: Gemini 3.8 Flash TTS (Google DeepMind); id `gemini-3.8-flash-tts` via Gemini API, https://generativelanguage.googleapis.com/v1beta/models/gemini-3.8-flash-tts:generateContent; $0.5 in / $9 out per 1M tokens; context 8,192 tokens; in: text; out: audio; status current; details https://postcutoff.com/m/gemini-3-8-flash-tts/
- Alternative: Gemini 3.8 Flash-Lite TTS (Google DeepMind); id `gemini-3.8-flash-lite-tts` via Gemini API, https://generativelanguage.googleapis.com/v1beta/models/gemini-3.8-flash-lite-tts:generateContent; $0.5 in / $6 out per 1M tokens; context 8,192 tokens; in: text; out: audio; status current; details https://postcutoff.com/m/gemini-3-8-flash-lite-tts/
- Open weights: Fish Audio Drama 3 (Fish Audio); id `drama-3` via Fish Audio API, https://api.fish.audio/v1/tts (model passed in `model` header; a missing or misspelled header falls back to s2.1-pro); per 1m utf8 bytes $2 (USD per 1M UTF-8 bytes on OpenRouter (2026-10-13). Fish Audio's own pricing page did not list Drama 3 yet; its press release says API usage is 80% off for the first 30 days, so this may be a launch price (list price unverified)); in: text, audio; out: audio; status current; details https://postcutoff.com/m/fish-audio-drama-3/
- Budget: Speechify Simba 3.2 (Speechify (SpeechifyAI)); no API id recorded; per 1m characters $10 (USD per 1M characters (entry tier); $6 at scale tier); in: text, audio; out: audio; status current; details https://postcutoff.com/m/speechify-simba-3-2/
- Why: Artificial Analysis TTS (Provider Voice Arena): Eleven v4 Turbo 1336 ($40/1M chars) and Eleven v4 1332 ($80), Qwen-Audio-3.1-TTS-Plus 1307 ($19.30), Cartesia Sonic 3.6 1292, Gemini 3.8 Flash TTS 1280 ($16.50), Inworld Realtime TTS-2 1270, Qwen-Audio-3.0-TTS-Plus 1266, Fish Audio Drama 3 Preview 1248 (open weights), Simba 3.2 1246 at $6.60. Both v4 variants are in elevenlabs-v4 (eleven_v4, eleven_v4_turbo). Qwen-Audio-3.1-TTS-Plus has no own registry file (API id unverified), so its 3.0 sibling is listed.
- Sources: https://artificialanalysis.ai/text-to-speech/leaderboard

### Speech-to-text and transcription [speech-to-text]
- Pick: Fun-Realtime-ASR-preview (Fun-ASR-Flash / Fun-ASR-Realtime) (Alibaba Cloud (Tongyi Lab)); id `fun-asr-realtime` via Alibaba Cloud Model Studio (streaming, WebSocket), wss://{WorkspaceId}.ap-southeast-1.maas.aliyuncs.com/api-ws/v1/inference; flash per second intl $0.000035, flash per second beijing $0.00003, realtime per second intl $0.00009, realtime per second beijing $0.000047 (USD per second of input audio, list prices excl. promotions (fun-asr-flash-2026-06-15 from its model page; fun-asr-realtime from the Model Studio pricing page). Artificial Analysis lists $2.82 per 1,000 min, which equals $0.000047/s (the Beijing fun-asr-realtime price)); in: audio; out: text; status current; details https://postcutoff.com/m/fun-realtime-asr/
- Alternative: Voxtral Small Transcribe 3 (Mistral AI); id `voxtral-small-transcribe-3-0` via Mistral API, https://api.mistral.ai/v1/audio/transcriptions; per minute $0.004 (USD per minute of audio (Mistral API)); context 32,000 tokens; in: audio; out: text; status preview; details https://postcutoff.com/m/voxtral-small-transcribe-3/
- Alternative: MAI-Transcribe-2 (Microsoft); id `MAI-Transcribe-2` via Azure Speech in Microsoft Foundry (Fast Transcription API, enhancedMode), https://{resource}.cognitiveservices.azure.com/speechtotext/transcriptions:transcribe?api-version=2025-10-15; per hour $0.1 (USD per hour of audio (limited-time promotional price through end of 2026; MAI-Transcribe-1.5 was $0.36/hr)); in: audio; out: text; status preview; details https://postcutoff.com/m/mai-transcribe-2/
- Alternative: Scribe v2 / Scribe v2 Medical (ElevenLabs); id `scribe_v2` via ElevenLabs API, https://api.elevenlabs.io/v1/speech-to-text; per hour $0.22 (per hour of audio (Scribe v2 and v2 Medical, API)); in: audio, video; out: text; status current; details https://postcutoff.com/m/elevenlabs-scribe-v2/
- Alternative: Voxtral Mini Transcribe 3 (+ Mini Transcribe Realtime 3) (Mistral AI); id `voxtral-mini-transcribe-3-0` via Mistral API (batch), https://api.mistral.ai/v1/audio/transcriptions; per minute batch $0.0035, per minute realtime $0.008 (USD per minute of audio (voxtral-mini-transcribe-3-0 batch / voxtral-mini-transcribe-realtime-3-0 streaming)); context 32,000 tokens; in: audio; out: text; status preview; details https://postcutoff.com/m/voxtral-mini-transcribe-3/
- Open weights: Voxtral Transcribe 2 (Mini Transcribe V2 + Voxtral Realtime) (Mistral AI); id `voxtral-mini-2602` via Mistral API (batch), https://api.mistral.ai/v1/audio/transcriptions; per minute batch $0.003, per minute realtime $0.006 (USD per minute of audio (voxtral-mini-2602 batch / voxtral-mini-transcribe-realtime-2602)); in: audio; out: text; status current; details https://postcutoff.com/m/voxtral-transcribe-2/
- Budget: Voxtral Mini Transcribe 3 (+ Mini Transcribe Realtime 3) (Mistral AI); id `voxtral-mini-transcribe-3-0` via Mistral API (batch), https://api.mistral.ai/v1/audio/transcriptions; per minute batch $0.0035, per minute realtime $0.008 (USD per minute of audio (voxtral-mini-transcribe-3-0 batch / voxtral-mini-transcribe-realtime-3-0 streaming)); context 32,000 tokens; in: audio; out: text; status preview; details https://postcutoff.com/m/voxtral-mini-transcribe-3/
- Why: Artificial Analysis speech-to-text (2026-10-25, word error rate): Alibaba Fun-ASR (realtime) #1 at 1.7%, Mistral Voxtral Small Transcribe 3 #3 at 2.1%, Voxtral Mini Transcribe 3 #5 at 2.2% ($0.0035/min). Which Fun-ASR id AA ranked is ambiguous (see the model page). MAI-Transcribe-2 remains strong but is in preview. Field note from a user's agent (unverified): for timing lyrics to a song, ElevenLabs Forced Alignment drifted on echoes and ad-libs, while Scribe v2 word timestamps (elevenlabs-scribe-v2, $0.22/hour) worked better.
- Sources: https://artificialanalysis.ai/speech-to-text , https://postcutoff.com/m/fun-realtime-asr/ , https://postcutoff.com/m/voxtral-small-transcribe-3/

### Real-time voice conversation (speech-to-speech agents) [realtime-voice]
- Pick: Grok Voice 3.0 (xAI); id `grok-voice-latest` via xAI API (Voice Agent, speech-to-speech); per minute $0.08 (USD per minute of audio ($4.80/hr), same as Grok Voice 2.0); in: audio, text; out: audio, text; status current; details https://postcutoff.com/m/grok-voice-3-0/
- Alternative: Qwen-Audio-3.1-Realtime (Plus) (Alibaba (Qwen)); id `qwen-audio-3.1-realtime-plus` via QwenCloud (Realtime WebSocket), wss://maas.qwencloudapi.com/api-ws/v1/realtime?model=qwen-audio-3.1-realtime-plus; audio input $6.4, text input $0.8, text output $6.4, audio output $24 (USD per 1M tokens (QwenCloud list price; audio/text output $24 when audio is generated, $6.40 text-only)); context 262,144 tokens; in: audio, text; out: audio, text; status current; details https://postcutoff.com/m/qwen-audio-3-1-realtime/
- Alternative: Gemini 3.8 Live (Google DeepMind); id `gemini-3.8-live` via Gemini Live API (WebSocket), wss://generativelanguage.googleapis.com/ws/google.ai.generativelanguage.v1beta.GenerativeService.BidiGenerateContent; $0.75 in / $4.5 out per 1M tokens; context 131,072 tokens; in: text, image, audio, video; out: text, audio; status current; details https://postcutoff.com/m/gemini-3-8-live/
- Alternative: Eleven Realtime v1 Max / Mini (ElevenLabs); id `eleven_realtime_v1_max` via ElevenLabs Realtime API (WebSocket, OpenAI Realtime-compatible), wss://api.elevenlabs.io/v1/convai/realtime; max input $1.8, max output $9, mini input $0.2, mini output $6, connected audio per minute $0.08 (USD per 1M tokens (input / output) plus $0.080 per additional minute of connected audio on the Realtime API (burst $0.160/min); same per-minute rate for Max and Mini); in: audio, text; out: audio, text; status current; details https://postcutoff.com/m/eleven-realtime-v1-max/
- Alternative: StepAudio 3 Realtime (StepFun); id `stepaudio-3-realtime-preview` via StepFun API (Realtime WebSocket); $0 in / $0 out per 1M tokens; in: audio, text; out: audio, text; status preview; details https://postcutoff.com/m/stepaudio-3-realtime/
- Alternative: GPT-Live-1.1 (OpenAI); id `not yet available (OpenAI: "Coming soon to the API"; API still serves gpt-live-1)` via OpenAI API; price not recorded (check the provider); in: audio, text; out: audio, text; status current; details https://postcutoff.com/m/gpt-live-1-1/
- Open weights: NVIDIA NemotronLabs VoiceChat 11B (and PersonaPlex-7B) (NVIDIA); id `nvidia/NVIDIA-NemotronLabs-VoiceChat-11B` via Hugging Face; price not recorded (check the provider); in: audio; out: audio, text; status current; details https://postcutoff.com/m/nemotron-voicechat/
- Why: Artificial Analysis speech-to-speech: Grok Voice 3.0 Big Bench Audio 98.5%, Full-Duplex Bench 97.0%, 0.81 s to first audio (Speech-to-Speech Index 86.9, top on Oct 13); StepAudio 3 Realtime scores highest (99.7% / 98.9%) but takes 8.83 s and is preview; Qwen Audio 3.0 Realtime Plus 99.2% / 98.4% at 1.54 s; Eleven Realtime v1 Max 0.57 s (not in registry). GPT-Live-1.1 powers ChatGPT voice but has no API yet ('coming soon'); use gpt-realtime-2-1 for OpenAI's API.
- Sources: https://artificialanalysis.ai/speech-to-speech , https://postcutoff.com/m/grok-voice-3-0/ , https://postcutoff.com/m/gpt-live-1-1/

### Music generation (songs with vocals, instrumentals) [music-generation]
- Pick: Suno v6 (v6, v6-wild, v6-mini) (Suno); id `v6` via Web app; free $0, pro per month $10, premier per month $30 (USD per month subscription (monthly billing; annual billing is 20% cheaper = $8 Pro / $24 Premier per month). Free: 50 credits/day, v6-mini only, no downloads, no commercial rights. Pro: 2,500 credits/month, 20 downloads/month, v6 + v6-wild, commercial rights. Premier: 10,000 credits/month, 60 downloads/month, Suno Studio.); in: text, audio, image, video; out: audio; status current; details https://postcutoff.com/m/suno-v6/
- Alternative: Mureka V9.5 (and O3) (Kunlun Tech (Skywork AI)); id `mureka-9.5` via Mureka API; price not recorded (check the provider); in: text, audio; out: audio; status current; details https://postcutoff.com/m/mureka-9-5/
- Alternative: Lyria 3.5 (Google DeepMind); id `lyria-3.5` via Gemini API (Interactions API), https://generativelanguage.googleapis.com/v1beta/interactions; per song $0.08 (per full song (~2 min). Lyria 3 Clip (30 s, lyria-3-clip-preview) $0.04 per clip; lyria-3-pro-preview $0.08 per song); in: text, image; out: audio; status current; details https://postcutoff.com/m/lyria-3-5/
- Alternative: Eleven Music v2.5 (ElevenLabs); id `music_v2_5` via ElevenLabs API, https://api.elevenlabs.io/v1/music; per minute $0.15 (per minute of generated music (API)); in: text, audio; out: audio; status current; details https://postcutoff.com/m/elevenlabs-music-v2-5/
- Open weights: MiniMax Music 3.0 (MiniMax); id `music-3.0` via MiniMax API (existing paying users only since 2026-08-20); per song $0.15 (USD per generation of up to 5 minutes (music-3.0 and music-2.6 API). Paid music APIs closed to NEW users from 2026-08-20; existing paying users keep access.); in: text; out: audio; status current; details https://postcutoff.com/m/minimax-music-3/
- Why: Artificial Analysis Music (vocals): Suno v6 1147, Suno v6-mini 1130, Mureka V9.5 1107, Lyria 3.5 1053, Eleven Music v2.5 1015, MiniMax Music 3.0 1000 (open weights). Suno v6 is the best and has no official API, but it works well through its web app (suno.com) for agents that can drive a browser (e.g. Claude in Chrome or other computer-use agents); creators in our video log commonly pair a language model with Suno this way. For a pure API pipeline use Mureka 9.5 (best via API) or Lyria 3.5 (GA in the Gemini API). Field notes from a user's agent (unverified, single runs): Suno m4a downloads carry the timed lyrics as a mov_text subtitle track (read it with ffmpeg); Suno v6 ignored a requested tempo (96 BPM asked, about 133 delivered) and a [key change] tag; the wand button next to the style field rewrites the style prompt toward the account's taste, so skip it when the exact prompt matters.
- Sources: https://artificialanalysis.ai/music/leaderboard/vocals , https://postcutoff.com/m/suno-v6/ , https://postcutoff.com/ai-video/

### Embeddings and retrieval (RAG, search) [embeddings]
- Pick: Gemini Embedding 2 (Google DeepMind); id `gemini-embedding-2` via Gemini API, https://generativelanguage.googleapis.com/v1beta/models/gemini-embedding-2:embedContent; input $0.2 (per 1M text input tokens; image $0.45/1M (~$0.00012 per image), audio $6.50/1M (~$0.00016/s), video $12.00/1M (~$0.00079 per frame)); context 8,192 tokens; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-embedding-2/
- Alternative: Cohere Embed 5 Pro (Cohere); id `embed-v5.0-pro` via Cohere API, https://api.cohere.com/v2/embed; input $0.12 (per 1M text tokens (images $0.40 per 1M tokens)); context 128,000 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/cohere-embed-v5-pro/
- Alternative: text-embedding-3-large (OpenAI); id `text-embedding-3-large` via OpenAI API, https://api.openai.com/v1/embeddings; input $0.13 (per 1M tokens (USD)); in: text; out: text; status current; details https://postcutoff.com/m/text-embedding-3-large/
- Open weights: EmbeddingGemma 2 (Google DeepMind); id `google/embeddinggemma-2` via Hugging Face; price not recorded (check the provider); context 8,192 tokens; in: text, image, video, audio; out: text; status current; details https://postcutoff.com/m/embeddinggemma-2/
- Budget: Cohere Embed 5 Fast (Cohere); id `embed-v5.0-fast` via Cohere API, https://api.cohere.com/v2/embed; input $0.08 (per 1M text tokens (images $0.40 per 1M tokens)); context 128,000 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/cohere-embed-v5-fast/
- Why: No current independent leaderboard checked (MTEB snapshots found online were from 2025). Picked on capability: Gemini Embedding 2 embeds text, images, video, audio and PDFs in one space ($0.20/1M); Cohere Embed 5 Pro is strong on visual documents (ViDoRe V3 85.8, vendor) and shares a vector space with Embed 5 Fast; EmbeddingGemma 2 is a 740M open multimodal model (MTEB code 78.68, vendor). Pair with cohere-rerank-4 for reranking.
- Sources: https://postcutoff.com/m/gemini-embedding-2/ , https://postcutoff.com/m/cohere-embed-v5-pro/ , https://postcutoff.com/m/embeddinggemma-2/

### Text translation and multilingual work [translation]
- Pick: Gemini 4 Argon (Google DeepMind); id `gemini-4-argon` via Gemini API, generateContent (paid tier only; works with google_search grounding and audio input, tested 2026-10-25; the Interactions API also works), https://generativelanguage.googleapis.com/v1beta/models/gemini-4-argon:generateContent; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-4-argon/
- Alternative: Claude Opus 5.5 (Anthropic); id `claude-opus-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $4 in / $20 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-opus-5-5/
- Alternative: GPT-6.1 Sol (OpenAI); id `gpt-6.1-sol` via OpenAI API, https://api.openai.com/v1/responses; $2 in / $10 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-1-sol/
- Alternative: Qwen3.8-Max (Alibaba (Qwen)); id `qwen3.8-max` via Alibaba Cloud Model Studio (DashScope, Singapore/Intl), https://{WorkspaceId}.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1; $2 in / $6 out per 1M tokens; context 1,000,000 tokens; in: text, image, video; out: text; status current; details https://postcutoff.com/m/qwen3-8-max/
- Open weights: Qwen3.8-Max (Alibaba (Qwen)); id `qwen3.8-max` via Alibaba Cloud Model Studio (DashScope, Singapore/Intl), https://{WorkspaceId}.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1; $2 in / $6 out per 1M tokens; context 1,000,000 tokens; in: text, image, video; out: text; status current; details https://postcutoff.com/m/qwen3-8-max/
- Budget: Gemini 3.8 Flash (Google DeepMind); id `gemini-3.8-flash` via Gemini API, https://generativelanguage.googleapis.com/v1beta/models/gemini-3.8-flash:generateContent; $0.75 in / $3.75 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-3-8-flash/
- Why: Unverified pick: we found no independent October 2026 translation ranking; official WMT26 results are expected after the conference on Oct 28-29, 2026. Choice is by overall capability and language coverage (Gemma/Gemini families cover 140+ languages; Qwen covers Chinese varieties well). Re-check once WMT26 publishes.
- Sources: https://artificialanalysis.ai/leaderboards/models

### Live speech-to-speech translation (interpreting) [speech-translation]
- Pick: Gemini 3.5 Live Translate (Google DeepMind); id `gemini-3.5-live-translate-preview` via Gemini Live API (WebSocket); audio input $3.5, audio output $21 (per 1M tokens (USD); ~$0.0053/min audio in, ~$0.0315/min audio out); context 131,072 tokens; in: audio; out: audio, text; status preview; details https://postcutoff.com/m/gemini-3-5-live-translate/
- Alternative: GPT-Realtime-Translate (OpenAI); id `gpt-realtime-translate` via OpenAI API, https://api.openai.com/v1/realtime/translations; per minute $0.034 (per minute of audio (USD), billed by duration not tokens); context 16,000 tokens; in: audio; out: audio, text; status current; details https://postcutoff.com/m/gpt-realtime-translate/
- Alternative: Qwen3.8-LiveTranslate (Flash Realtime) (Alibaba (Qwen)); id `qwen3.8-livetranslate-flash-realtime` via QwenCloud (Realtime WebSocket), wss://maas.qwencloudapi.com/api-ws/v1/realtime?model=qwen3.8-livetranslate-flash-realtime; audio input $7.5, image input $0.55, text output $20, audio output $30 (USD per 1M tokens (QwenCloud list price; press estimates about $1.54 per hour of speech in and out)); context 53,248 tokens; in: audio, image, text; out: audio, text; status current; details https://postcutoff.com/m/qwen3-8-livetranslate/
- Why: Vendor specs only: Gemini 3.5 Live Translate (preview) translates audio-to-audio in 70+ languages / 2,000+ pairs and keeps the speaker's voice; GPT-Realtime-Translate covers 70+ input and 13 output languages; Qwen3.8-LiveTranslate understands 60 and speaks 29 languages with per-speaker voice cloning. No independent comparison found.
- Sources: https://postcutoff.com/m/gemini-3-5-live-translate/ , https://postcutoff.com/m/gpt-realtime-translate/ , https://postcutoff.com/m/qwen3-8-livetranslate/

### Fast and cheap high-volume work (classification, extraction, routing, subagents) [fast-cheap]
- Pick: Claude Haiku 5.5 (Anthropic); id `claude-haiku-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $0.1 in / $0.5 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image; out: text; status current; details https://postcutoff.com/m/claude-haiku-5-5/
- Alternative: DeepSeek-V4.1-Flash (DeepSeek); id `deepseek-flash` via DeepSeek API, https://api.deepseek.com; $0.3 in / $1.2 out per 1M tokens; context 1,000,000 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/deepseek-v4-1-flash/
- Alternative: MiniMax-M3.1-Flash (MiniMax); id `MiniMax-M3.1-Flash` via MiniMax API, https://api.minimax.io/anthropic; $0.15 in / $0.6 out per 1M tokens; context 1,000,000 tokens; in: text, image, video; out: text; status current; details https://postcutoff.com/m/minimax-m3-1-flash/
- Alternative: GPT-6 Luna (OpenAI); id `gpt-6-luna` via OpenAI API, https://api.openai.com/v1/responses; $0.1 in / $0.5 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-05; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-luna/
- Alternative: Gemini 3.5 Flash-Lite (Google DeepMind); id `gemini-3.5-flash-lite` via Gemini API, https://generativelanguage.googleapis.com/v1beta/models/gemini-3.5-flash-lite:generateContent; $0.3 in / $2.5 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-3-5-flash-lite/
- Open weights: DeepSeek-V4.1-Flash (DeepSeek); id `deepseek-flash` via DeepSeek API, https://api.deepseek.com; $0.3 in / $1.2 out per 1M tokens; context 1,000,000 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/deepseek-v4-1-flash/
- Budget: GPT-6 Luna (OpenAI); id `gpt-6-luna` via OpenAI API, https://api.openai.com/v1/responses; $0.1 in / $0.5 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-05; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-luna/
- Why: Artificial Analysis: Claude Haiku 5.5 (max) 43 at $0.21 per index task and 234 tok/s; DeepSeek V4.1 Flash 45 at $0.26 and 169 tok/s; MiniMax-M3.1-Flash 43 at $0.22; GPT-6 Luna 38 at $0.07 (cheapest); Gemini 3.5 Flash-Lite 22 but 414 tok/s. For raw speed: Mercury 2.5 (714 tok/s on AA) and mercury-3 (diffusion, vendor 1,068 tok/s). Haiku and Luna both cost $0.10/$0.50 per 1M at the low tier.
- Sources: https://artificialanalysis.ai/leaderboards/models , https://postcutoff.com/m/claude-haiku-5-5/

### Best open-weights model overall (self-hosting) [open-weights]
- Pick: MiMo-V2.6-Pro (Xiaomi); id `mimo-v2.6-pro` via Xiaomi MiMo API; $0.435 in / $0.87 out per 1M tokens; context 1,000,000 tokens; in: text, image, video, audio; out: text; status current; details https://postcutoff.com/m/mimo-v2-6-pro/
- Alternative: DeepSeek-V4.1-Flash (DeepSeek); id `deepseek-flash` via DeepSeek API, https://api.deepseek.com; $0.3 in / $1.2 out per 1M tokens; context 1,000,000 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/deepseek-v4-1-flash/
- Alternative: GLM-5.3 (Zhipu AI (Z.ai)); id `glm-5.3` via Z.ai API, https://api.z.ai/api/paas/v4; $1.4 in / $4.4 out per 1M tokens; context 1,000,000 tokens; in: text; out: text; status current; details https://postcutoff.com/m/glm-5-3/
- Alternative: Kimi K3 (Moonshot AI); id `kimi-k3` via Kimi API (Moonshot), https://api.moonshot.ai/v1; $3 in / $15 out per 1M tokens; context 1,048,576 tokens; in: text, image, video; out: text; status current; details https://postcutoff.com/m/kimi-k3/
- Open weights: MiMo-V2.6-Pro (Xiaomi); id `mimo-v2.6-pro` via Xiaomi MiMo API; $0.435 in / $0.87 out per 1M tokens; context 1,000,000 tokens; in: text, image, video, audio; out: text; status current; details https://postcutoff.com/m/mimo-v2-6-pro/
- Budget: GLM-5.3-Flash / FlashX (Zhipu AI (Z.ai)); id `glm-5.3-flash` via Z.ai API, https://api.z.ai/api/paas/v4; $0.15 in / $0.5 out per 1M tokens; context 1,000,000 tokens; in: text, image, video, pdf; out: text; status current; details https://postcutoff.com/m/glm-5-3-flash/
- Why: Artificial Analysis open-weights list: MiMo-V2.6-Pro 46 (MIT, 1.02T/42B active), DeepSeek V4.1 Flash 1017 45, GLM-5.3 45, Step 5 Preview 44, GLM-5.3-Flash 42, Muse Spark Open 39. Kimi K3.1 Pro (55) would lead by 9 points once its weights ship (promised by Nov 4, 2026); Mistral Large 4 weights are promised by end of October.
- Sources: https://artificialanalysis.ai/models/open-source , https://postcutoff.com/e/2026-10-23-kimi-k3-1-pro/

### On-device and local models (phones, laptops, edge) [on-device]
- Pick: Gemma 4 (Google DeepMind); id `google/gemma-4-31B-it` via Hugging Face; $0.09 in / $0.34 out per 1M tokens; context 262,144 tokens; in: text, image, audio, video; out: text; status current; details https://postcutoff.com/m/gemma-4/
- Alternative: LFM2.5-24B-A2B (Liquid AI); no API id recorded; $0.1 in / $1 out per 1M tokens; context 131,072 tokens; in: text; out: text; status current; details https://postcutoff.com/m/liquid-lfm2-5-24b-a2b/
- Alternative: Qwen3.8-27B (Alibaba (Qwen)); id `qwen/qwen3.8-27b` via OpenRouter; $0.2 in / $2.5 out per 1M tokens; context 262,144 tokens; in: text, image, video; out: text; status current; details https://postcutoff.com/m/qwen3-8-27b/
- Alternative: Phi-4-Reasoning-Vision-15B (Microsoft); no API id recorded; price not recorded (check the provider); context 16,384 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/phi-4-reasoning-vision-15b/
- Open weights: Gemma 4 (Google DeepMind); id `google/gemma-4-31B-it` via Hugging Face; $0.09 in / $0.34 out per 1M tokens; context 262,144 tokens; in: text, image, audio, video; out: text; status current; details https://postcutoff.com/m/gemma-4/
- Budget: Liquid AI d1-3B (Liquid AI); no API id recorded; price not recorded (check the provider); context 32,768 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/liquid-d1-3b/
- Why: No independent on-device ranking checked. Gemma 4 (Apache 2.0) ships E2B/E4B edge sizes with native audio and vision plus 12B-31B sizes; Qwen3.8 27B scores 34 on AA's open list; Liquid d1 3B answers typed decisions in 8-50 ms. For local embeddings use embeddinggemma-2, for local TTS kokoro-82m.
- Sources: https://postcutoff.com/m/gemma-4/ , https://artificialanalysis.ai/models/open-source , https://postcutoff.com/m/liquid-d1-3b/

### Computer use and browser agents [computer-use]
- Pick: Claude Opus 5.5 (Anthropic); id `claude-opus-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $4 in / $20 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-opus-5-5/
- Alternative: Claude Fable 5.1 (Anthropic); id `claude-fable-5-1` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $10 in / $50 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-fable-5-1/
- Alternative: GPT-6 Astra (OpenAI); id `gpt-6-astra` via OpenAI API, https://api.openai.com/v1/responses; $10 in / $50 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-astra/
- Alternative: GPT-6.1 Sol (OpenAI); id `gpt-6.1-sol` via OpenAI API, https://api.openai.com/v1/responses; $2 in / $10 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-1-sol/
- Alternative: Gemini 4 Argon (Google DeepMind); id `gemini-4-argon` via Gemini API, generateContent (paid tier only; works with google_search grounding and audio input, tested 2026-10-25; the Interactions API also works), https://generativelanguage.googleapis.com/v1beta/models/gemini-4-argon:generateContent; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-4-argon/
- Budget: Claude Haiku 5.5 (Anthropic); id `claude-haiku-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $0.1 in / $0.5 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image; out: text; status current; details https://postcutoff.com/m/claude-haiku-5-5/
- Why: Benchmarks disagree by version: Meta's table has OSWorld V2.1 (partial) Opus 5.5 81.8, Fable 5.1 80.7, Muse Wonder Full 78.4, GPT-6 Astra 75.8; Google's table has OSWorld-2.0 offline subset Astra 72.6 vs Argon 69.2, and Agent's Last Exam Argon 39.5 vs Opus 5.5 38.2 vs Astra 34.2. GPT-6.1 Sol is within 2.1 points of Astra on OSWorld 2.0 at ~1/7 the cost per task (OpenAI). Haiku 5.5 reports OSWorld 2.1 72.4% at $0.10/$0.50. No verified open-weights leader.
- Sources: https://postcutoff.com/e/2026-10-12-meta-muse-wonder/ , https://deepmind.google/models/model-cards/gemini-4-argon/ , https://postcutoff.com/m/gpt-6-1-sol/ , https://postcutoff.com/m/claude-haiku-5-5/

### Web search and grounded answers (research with citations) [web-search]
- Pick: GPT-6.1 Sol (OpenAI); id `gpt-6.1-sol` via OpenAI API, https://api.openai.com/v1/responses; $2 in / $10 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-1-sol/
- Alternative: Gemini 3.8 Flash (Google DeepMind); id `gemini-3.8-flash` via Gemini API, https://generativelanguage.googleapis.com/v1beta/models/gemini-3.8-flash:generateContent; $0.75 in / $3.75 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-3-8-flash/
- Alternative: Muse Wonder (Meta); id `muse-wonder-1.0` via Meta Model API, https://api.meta.ai/v1/chat/completions; $4 in / $20 out per 1M tokens; context 1,048,576 tokens; in: text, image, video, audio, pdf; out: text; status preview; details https://postcutoff.com/m/muse-wonder/
- Alternative: Gemini Deep Research agent (API) (Google); id `deep-research-preview-04-2026` via Gemini API (Interactions API), https://generativelanguage.googleapis.com/v1beta/interactions; per task estimate low $1, per task estimate high $3, max per task estimate low $3, max per task estimate high $7 (USD per research task, Google's own estimate (Deep Research ~80 searches, ~250k input / ~60k output tokens; Max up to ~160 searches, ~900k in / ~80k out). Billed pay-as-you-go at the underlying Gemini model's rates plus tool fees; 'estimates based on preview rates and subject to change'); in: text; out: text; status preview; details https://postcutoff.com/m/gemini-deep-research-agent/
- Budget: Gemini 3.8 Flash (Google DeepMind); id `gemini-3.8-flash` via Gemini API, https://generativelanguage.googleapis.com/v1beta/models/gemini-3.8-flash:generateContent; $0.75 in / $3.75 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-3-8-flash/
- Why: No current search leaderboard read (Arena Search returned 403). Picked on hosted tooling and factuality: GPT-6.1 Sol has built-in web search plus multi-agent delegation in the Responses API; Gemini 3.8 Flash has Google Search and Maps grounding; Muse Wonder has web search grounding at $2.50 per 1,000 queries; for multi-minute report-style research use Google's Deep Research agent (preview). On hallucination, AA measured Gemini 4 Argon at 15% on AA-Omniscience vs 51% for GPT-6 Astra (Sept 30 run), but Argon's grounding support is not documented in our registry.
- Sources: https://postcutoff.com/m/gpt-6-1-sol/ , https://postcutoff.com/m/gemini-3-8-flash/ , https://postcutoff.com/m/gemini-4-argon/ , https://postcutoff.com/m/gemini-deep-research-agent/

### Creative writing: fiction, lyrics, scripts, poetry [creative-writing]
- Pick: Claude Fable 5.1 (Anthropic); id `claude-fable-5-1` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $10 in / $50 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-fable-5-1/
- Alternative: GPT-6 Astra (OpenAI); id `gpt-6-astra` via OpenAI API, https://api.openai.com/v1/responses; $10 in / $50 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-astra/
- Alternative: Claude Opus 5.5 (Anthropic); id `claude-opus-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $4 in / $20 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-opus-5-5/
- Alternative: GPT-6 Sol (OpenAI); id `gpt-6-sol` via OpenAI API, https://api.openai.com/v1/responses; $2 in / $10 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-sol/
- Alternative: Kimi K3.1 Pro (Moonshot AI); id `kimi-k3.1-pro` via Kimi API (Moonshot), https://api.moonshot.ai/v1; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/kimi-k3-1-pro/
- Alternative: Grok 4.7 (xAI); id `grok-4.7` via xAI API, https://api.x.ai/v1/chat/completions; $2 in / $6 out per 1M tokens; context 500,000 tokens; knows up to 2026-05; in: text, image; out: text; status current; details https://postcutoff.com/m/grok-4-7/
- Alternative: Gemini 4 Argon (Google DeepMind); id `gemini-4-argon` via Gemini API, generateContent (paid tier only; works with google_search grounding and audio input, tested 2026-10-25; the Interactions API also works), https://generativelanguage.googleapis.com/v1beta/models/gemini-4-argon:generateContent; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-4-argon/
- Open weights: GLM-5.3 (Zhipu AI (Z.ai)); id `glm-5.3` via Z.ai API, https://api.z.ai/api/paas/v4; $1.4 in / $4.4 out per 1M tokens; context 1,000,000 tokens; in: text; out: text; status current; details https://postcutoff.com/m/glm-5-3/
- Budget: GPT-6 Sol (OpenAI); id `gpt-6-sol` via OpenAI API, https://api.openai.com/v1/responses; $2 in / $10 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-sol/
- Why: Two independent LLM-judged boards, read 2026-10-25. EQ-Bench Creative Writing v3 (Elo): GPT-6 Astra #1 2174.0, Claude Fable 5.1 #2 2162.7, Claude Opus 5 (legacy) 2133.5, GPT-6 Sol 2124.6, Kimi K3 2083.5, GLM-5.3 2076.7, Claude Opus 5.5 2050.8, Grok 4.7 2012.2 (highest rubric score, 17.16), Gemini 4 Argon only #31 at 1758.0. Lech Mazur's Short-Story benchmark (lechmazur/writing, 600-800-word stories with 10 required elements, pairwise judged; last update Oct 23, 2026): Fable 5.1 (high) #1 3.745, Opus 5.5 (high) 3.693, Opus 5 (xhigh, legacy) 3.685, GPT-6 Astra (high) #4 3.416 (3rd among current models), GLM-5.3 (max) 2.950, Kimi K3.1 Pro 2.723, Gemini 4 Argon 0.719 (#20), Grok 4.7 0.351 (#26). Fable 5.1 is top-2 on both; best OpenAI is GPT-6 Astra ($10/$50), with GPT-6 Sol ($2/$10) close on EQ-Bench; GPT-6.1 Sol, Sonnet 5.5 and Haiku 5.5 are on neither board. Open: GLM-5.3 ranks above Kimi K3 on the short-story board. EQ-Bench Longform was not used: its table has no model newer than Claude Sonnet 4.6 (Feb 2026). Arena's creative-writing category returned HTTP 403; Artificial Analysis has no creative-writing index. Song lyrics: Suno and other music models can write lyrics themselves, but a strong LLM usually writes better ones; write them here, then pass them to suno-v6 (see music-generation).
- Sources: https://eqbench.com/creative_writing.html , https://eqbench.com/creative_writing_longform.html , https://github.com/lechmazur/writing

### Deep research reports with citations (multi-step web research) [deep-research]
- Pick: Gemini Deep Research agent (API) (Google); id `deep-research-preview-04-2026` via Gemini API (Interactions API), https://generativelanguage.googleapis.com/v1beta/interactions; per task estimate low $1, per task estimate high $3, max per task estimate low $3, max per task estimate high $7 (USD per research task, Google's own estimate (Deep Research ~80 searches, ~250k input / ~60k output tokens; Max up to ~160 searches, ~900k in / ~80k out). Billed pay-as-you-go at the underlying Gemini model's rates plus tool fees; 'estimates based on preview rates and subject to change'); in: text; out: text; status preview; details https://postcutoff.com/m/gemini-deep-research-agent/
- Alternative: GPT-6 Astra (OpenAI); id `gpt-6-astra` via OpenAI API, https://api.openai.com/v1/responses; $10 in / $50 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-astra/
- Alternative: Claude Opus 5.5 (Anthropic); id `claude-opus-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $4 in / $20 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-opus-5-5/
- Alternative: Kimi K3.1 Pro (Moonshot AI); id `kimi-k3.1-pro` via Kimi API (Moonshot), https://api.moonshot.ai/v1; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/kimi-k3-1-pro/
- Alternative: GPT-6.1 Sol (OpenAI); id `gpt-6.1-sol` via OpenAI API, https://api.openai.com/v1/responses; $2 in / $10 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-1-sol/
- Alternative: Gemini 4 Argon (Google DeepMind); id `gemini-4-argon` via Gemini API, generateContent (paid tier only; works with google_search grounding and audio input, tested 2026-10-25; the Interactions API also works), https://generativelanguage.googleapis.com/v1beta/models/gemini-4-argon:generateContent; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-4-argon/
- Alternative: Grok 4.7 (xAI); id `grok-4.7` via xAI API, https://api.x.ai/v1/chat/completions; $2 in / $6 out per 1M tokens; context 500,000 tokens; knows up to 2026-05; in: text, image; out: text; status current; details https://postcutoff.com/m/grok-4-7/
- Open weights: Ling 3.1 Flash (inclusionAI (Ant Group)); id `inclusionai/ling-3.1-flash` via OpenRouter; $0 in / $0 out per 1M tokens; context 1,000,000 tokens; in: text; out: text; status current; details https://postcutoff.com/m/ling-3-1-flash/
- Budget: GPT-6.1 Sol (OpenAI); id `gpt-6.1-sol` via OpenAI API, https://api.openai.com/v1/responses; $2 in / $10 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-1-sol/
- Why: No independent October 2026 ranking of deep-research agents found (FutureSearch's Deep Research Bench page was last updated June 25, 2026 and shows results only as an image; the DeepResearch Bench Space did not load). Picked on capability: Google's Deep Research agent is the only dedicated report-writing research agent with an API in the registry (Interactions API, background runs, a Max tier; preview). For a model running its own search loop, BrowseComp as compiled by BenchLM (2026-10-23, mostly vendor-reported): Atria Dawn Preview 92.5% (not in registry), Kimi K3.1 Pro 92.1%, Ling 3.1 Flash 91.7% (open, MIT), GPT-6 Astra 91.5%, Kimi K3 91.2%; no BrowseComp figure found for Claude Opus 5.5, GPT-6.1 Sol, Gemini 4 Argon or Grok 4.7. For long report-style deliverables, AA's GDPval-AA (agentic, with web browsing) has Opus 5.5 first (1866 Elo). GPT-6.1 Sol has hosted web search and multi-agent delegation in the Responses API at $2/$10. ChatGPT deep research and Claude Research are app features, not API models, and are not in the registry.
- Sources: https://futuresearch.ai/deep-research-bench , https://benchlm.ai/md/benchmarks/browsecomp.md , https://artificialanalysis.ai/evaluations/gdpval-aa , https://postcutoff.com/m/gemini-deep-research-agent/ , https://postcutoff.com/m/gpt-6-1-sol/

### Tool use and long agent runs (function calling, MCP, multi-step tasks) [agentic-tool-use]
- Pick: Claude Opus 5.5 (Anthropic); id `claude-opus-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $4 in / $20 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-opus-5-5/
- Alternative: Kimi K3.1 Pro (Moonshot AI); id `kimi-k3.1-pro` via Kimi API (Moonshot), https://api.moonshot.ai/v1; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/kimi-k3-1-pro/
- Alternative: Claude Sonnet 5.5 (Anthropic); id `claude-sonnet-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $2 in / $10 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image, pdf; out: text; status current; details https://postcutoff.com/m/claude-sonnet-5-5/
- Alternative: Gemini 4 Argon (Google DeepMind); id `gemini-4-argon` via Gemini API, generateContent (paid tier only; works with google_search grounding and audio input, tested 2026-10-25; the Interactions API also works), https://generativelanguage.googleapis.com/v1beta/models/gemini-4-argon:generateContent; $2 in / $10 out per 1M tokens; context 1,048,576 tokens; knows up to 2026-03; in: text, image, audio, video, pdf; out: text; status current; details https://postcutoff.com/m/gemini-4-argon/
- Alternative: Grok 4.7 (xAI); id `grok-4.7` via xAI API, https://api.x.ai/v1/chat/completions; $2 in / $6 out per 1M tokens; context 500,000 tokens; knows up to 2026-05; in: text, image; out: text; status current; details https://postcutoff.com/m/grok-4-7/
- Alternative: GPT-6.1 Sol (OpenAI); id `gpt-6.1-sol` via OpenAI API, https://api.openai.com/v1/responses; $2 in / $10 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-1-sol/
- Alternative: GPT-6 Astra (OpenAI); id `gpt-6-astra` via OpenAI API, https://api.openai.com/v1/responses; $10 in / $50 out per 1M tokens; context 1,050,000 tokens; knows up to 2026-04; in: text, image; out: text; status current; details https://postcutoff.com/m/gpt-6-astra/
- Open weights: DeepSeek-V4.1-Flash (DeepSeek); id `deepseek-flash` via DeepSeek API, https://api.deepseek.com; $0.3 in / $1.2 out per 1M tokens; context 1,000,000 tokens; in: text, image; out: text; status current; details https://postcutoff.com/m/deepseek-v4-1-flash/
- Budget: Claude Haiku 5.5 (Anthropic); id `claude-haiku-5-5` via Anthropic API (Claude API), https://api.anthropic.com/v1/messages; $0.1 in / $0.5 out per 1M tokens; context 1,000,000 tokens; knows up to 2026-06; in: text, image; out: text; status current; details https://postcutoff.com/m/claude-haiku-5-5/
- Why: Artificial Analysis GDPval-AA v2.1 (220 real work tasks, shell plus web browsing in an agent loop, Elo, read 2026-10-25): Claude Opus 5.5 (max) #1 1866, Muse Wonder Full 1857 (no API price), Kimi K3.1 Pro 1846, Claude Sonnet 5.5 1838, DeepSeek V4.1 Flash 1017 (max) 1746 (open weights), Grok 4.7 (xhigh) 1714, MiMo-V2.6-Pro 1683, Gemini 4 Argon (high) 1624, Claude Haiku 5.5 (max) 1619, GPT-6.1 Sol (max) 1596, GPT-6 Astra (max) 1574. Terminal-Bench 4.0: Opus 5.5 66.4% (vendor), Kimi K3.1 Pro 67% (AA's highest), GPT-6 Astra 58.2%, Argon 57.4%. Google's own table has Argon ahead on AutomationBench (51.3% vs Opus 5.5 42.5%, Astra 41.4%) and Vals Finance Agent v2 (65.4% vs 58.6%). Best OpenAI here is GPT-6.1 Sol, not Astra. AA's tau2-bench Telecom and Terminal-Bench Hard pages show only older models in their text, so no current tau-bench numbers are quoted.
- Sources: https://artificialanalysis.ai/evaluations/gdpval-aa , https://artificialanalysis.ai/evaluations/tau2-bench , https://deepmind.google/models/model-cards/gemini-4-argon/ , https://postcutoff.com/e/2026-10-23-kimi-k3-1-pro/ , https://postcutoff.com/e/2026-09-22-claude-opus-5-5/