As of: 2026-11-13 13:49 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/m/glm-5-3-flash/

# GLM-5.3-Flash / FlashX

Zhipu AI (Z.ai), GLM-5 family. Status: current. Type: multimodal model. Released: 2026-08. Training cutoff: not published.

## How to call it

| Provider | Model id | Endpoint | Docs |
|---|---|---|---|
| Z.ai API | `glm-5.3-flash` | https://api.z.ai/api/paas/v4 | https://docs.z.ai/guides/vlm/glm-5.3-flash |
| Z.ai API (fast) | `glm-5.3-flashx` | https://api.z.ai/api/paas/v4 | https://docs.z.ai/guides/vlm/glm-5.3-flash |
| OpenRouter | `z-ai/glm-5.3-flash` |  | https://openrouter.ai/z-ai/glm-5.3-flash |
| OpenRouter (FlashX) | `z-ai/glm-5.3-flashx` |  | https://openrouter.ai/z-ai/glm-5.3-flashx |
| RunInfra (third-party inference, also via Vercel AI Gateway) (Vendor claims (2026-09-30): rewritten FP8 kernels, ~670 tok/s on Vercel AI Gateway, $0.11 in / $0.45 out / $0.03 cached per 1M tokens, 1M context, runs on NVIDIA and AMD, OpenAI- and Anthropic-compatible APIs, zero data retention. Not independently benchmarked.) |  |  | https://x.com/runinfrai/status/2105186912634057086 https://runinfra.ai/inference-api/glm-5-3-flash |
| Hugging Face |  |  | https://huggingface.co/zai-org/GLM-5.3-Flash |
| Web app |  |  | https://chat.z.ai |

## Pricing

- Input: $0.15 per 1M tokens
- Output: $0.50 per 1M tokens
- Cache read: $0.03 per 1M tokens
- Unit as published: per 1M tokens (USD) for glm-5.3-flash; glm-5.3-flashx (~200 tok/s): 0.37 in / 1.25 out / 0.075 cached
- Source: https://docs.z.ai/guides/overview/pricing
- Last checked: 2026-09-30

## Specs

- Input: text, image, video, pdf
- Output: text
- Context window: 1,000,000 tokens
- Max output: 128,000 tokens
- Open weights: yes
- Licence: mit

## What it has not seen

- No published training cutoff. Counted from its release (2026-08): at least 1,604 AI events since (356 major, 72 historic), as of 2026-11-13.
- Everything since July 2026: https://postcutoff.com/since/2026-07/

## What stands out

- First native multimodal GLM-5 model: First GLM-5-series model with native vision (image, video, file input); vision used inside the coding loop (UI replication, Blender, browser/computer-use agents). Source: https://docs.z.ai/guides/vlm/glm-5.3-flash
- Sparse + linear attention hybrid (first): 320B total / 18B active; Z.ai claims it is the first open-source frontier model combining sparse and linear attention (3.01x less attention compute, 4.44x smaller KV cache vs GLM-5.3). Source: https://docs.z.ai/guides/vlm/glm-5.3-flash
- Office deliverables with visual self-check: Produces PPTX/PDF/DOCX/XLSX and renders them to catch overflow and layout issues. Source: https://docs.z.ai/guides/vlm/glm-5.3-flash

## Notes

Z.ai says it beats GLM-5.2 at a fraction of the cost; 3x Coding Plan quota vs GLM-5.3 (FlashX not yet on the plan). Thinking cannot be disabled. 'first' claim is the vendor's own. ARC Prize verified (Sept 29 2026): ARC-AGI-2 65.8% at $0.09/task, ARC-AGI-1 91.0% at $0.04/task; it was tested earlier in stealth on OpenRouter as Ox Alpha (x.com/arcprize/status/2104996180425884119).

Cheap multimodal GLM for visual coding, agents and office documents; open weights under MIT.

```bash
curl https://api.z.ai/api/paas/v4/chat/completions \
 -H "Authorization: Bearer $ZAI_API_KEY" -H "Content-Type: application/json" \
 -d '{"model":"glm-5.3-flash","messages":[{"role":"user","content":[{"type":"image_url","image_url":{"url":"https://example.com/ui.png"}},{"type":"text","text":"Rebuild this UI in React"}]}]}'
```

Sources: https://docs.z.ai/guides/vlm/glm-5.3-flash · https://docs.z.ai/guides/overview/pricing · https://huggingface.co/zai-org/GLM-5.3-Flash

Model page: https://postcutoff.com/m/glm-5-3-flash/