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/kimi-k3/

# Kimi K3

Moonshot AI. Status: current. Type: reasoning model. Released: 2026-07-16. Training cutoff: not published.

## How to call it

| Provider | Model id | Endpoint | Docs |
|---|---|---|---|
| Kimi API (Moonshot) | `kimi-k3` | https://api.moonshot.ai/v1 | https://platform.kimi.ai/docs/guide/kimi-k3-quickstart |
| Alibaba Cloud Model Studio | `kimi-k3` |  | https://www.alibabacloud.com/help/en/model-studio/models |
| OpenRouter | `moonshotai/kimi-k3` |  | https://openrouter.ai/moonshotai/kimi-k3 |
| Hugging Face |  |  | https://huggingface.co/moonshotai/Kimi-K3 |
| Web app |  |  | https://www.kimi.com |

## Pricing

- Input: $3 per 1M tokens
- Output: $15 per 1M tokens
- Cache read: $0.30 per 1M tokens
- Cache write, 5 minutes: $3 per 1M tokens
- Cache write, 1 hour: $6 per 1M tokens
- Unit as published: per 1M tokens (USD)
- Source: https://platform.kimi.ai/docs/pricing/chat
- Last checked: 2026-09-29

## Specs

- Input: text, image, video
- Output: text
- Context window: 1,048,576 tokens
- Max output: not published
- Open weights: yes
- Licence: kimi-k3 (custom)

## What it has not seen

- No published training cutoff. Counted from its release (2026-07-16): at least 1,760 AI events since (401 major, 80 historic), as of 2026-11-13.
- Everything since June 2026: https://postcutoff.com/since/2026-06/

## What stands out

- First open 3T-class model (first): 2.8T-parameter MoE (16 of 896 experts active) - Moonshot's claim: the first open model at this scale; weights released after launch (promised by 2026-07-27). Source: https://www.kimi.com/blog/kimi-k3
- Kimi Delta Attention + Attention Residuals: Hybrid linear attention (KDA) and AttnRes; ~2.5x the scaling efficiency of K2 per Moonshot. Source: https://platform.kimi.ai/docs/guide/kimi-k3-quickstart
- Native vision with 1M context: Native visual understanding (image and video) and a 1,048,576-token window; strong at coding tasks that use screenshots/visual feedback (games, frontend, CAD). Source: https://platform.kimi.ai/docs/guide/kimi-k3-quickstart
- Always-on thinking with effort control: Thinking cannot be disabled; reasoning_effort low/high/max (default max). Source: https://platform.kimi.ai/docs/guide/kimi-k3-quickstart

## Notes

Moonshot flagship. API unlocked after a minimum $1 top-up. Chat Completions, Responses and Anthropic-compatible Messages supported. Max output and knowledge cutoff not verified. Docs moved to platform.kimi.ai (platform.moonshot.ai still serves).

Moonshot's frontier open model for long-horizon coding, knowledge work and deep reasoning; 1M context, vision.

```bash
curl https://api.moonshot.ai/v1/chat/completions \
 -H "Authorization: Bearer $MOONSHOT_API_KEY" -H "Content-Type: application/json" \
 -d '{"model":"kimi-k3","reasoning_effort":"high","messages":[{"role":"user","content":"Hello"}]}'
```

Sources: https://platform.kimi.ai/docs/guide/kimi-k3-quickstart · https://platform.kimi.ai/docs/pricing/chat · https://platform.kimi.ai/docs/models · https://www.kimi.com/blog/kimi-k3

Model page: https://postcutoff.com/m/kimi-k3/