Kimi K3
Canonical page on the main site: chinaaihub.com/models/kimi-k3
Provider
Model Family
Kimi K-series
Release Date
2026-07-16
Status
active
Architecture
MoE: 2.8T total / 104B activated; 93 layers (1 dense); 896 experts (16 selected + 2 shared per token); 69 KDA + 24 Gated MLA layers; hidden dim 7168; SiTU-GLU; MoonViT-V2 vision encoder (401M); MXFP4 weights / MXFP8 activations
Parameter Information
- total_parameters: 2.8T
- active_parameters: 104B
Context Window
1048576
Maximum Output
1048576
Capabilities
- reasoning: Yes
- coding: Yes
- vision: Yes
- video: Yes
- tool_calling: Yes
- structured_output: Yes
- agent_capability: Yes
Open Weight
Yes
License
Kimi K3 License (permissive MIT-style, but Model-as-a-Service operators with >$20M aggregate revenue over any 12 months must sign a separate agreement; products with >100M MAU or >$20M monthly revenue must display 'Kimi K3' in the UI)
Self Hosting
Yes
Api Available
Yes
Pricing
- input_price_per_1m: 3.0
- output_price_per_1m: 15.0
- currency: USD
- pricing_ref: moonshot-ai
Official Api
Yes
Cloud Providers
- Moonshot AI Platform
Benchmark Results
| benchmark | score | metric | date | source_type | source_url |
|---|---|---|---|---|---|
| GPQA Diamond | 93.5 | accuracy | 2026-07 | vendor_reported | https://github.com/MoonshotAI/Kimi-K3 |
| HLE-Full | 43.5 (56.0 with tools) | accuracy | 2026-07 | vendor_reported | https://github.com/MoonshotAI/Kimi-K3 |
| DeepSWE | 67.5 | accuracy | 2026-07 | vendor_reported | https://github.com/MoonshotAI/Kimi-K3 |
| Terminal-Bench 2.1 | 88.3 | accuracy | 2026-07 | vendor_reported | https://github.com/MoonshotAI/Kimi-K3 |
| MMMU-Pro | 81.6 | accuracy | 2026-07 | vendor_reported | https://github.com/MoonshotAI/Kimi-K3 |
| Video-MME (with subtitles) | 90.0 | accuracy | 2026-07 | vendor_reported | https://github.com/MoonshotAI/Kimi-K3 |
Known Limitations
- Temperature fixed at 1.0 and top_p at 0.95 - cannot be modified
- API access requires a minimum $1 top-up
- Modality documentation inconsistent: architecture table says Text+Image, while the README, launch blog and API guide also show video input
- Benchmarks vendor-reported; some comparison scores cited from Artificial Analysis
Last Verified
2026-09-20
Sources
| source_name | source_url | source_type | last_verified | confidence | published_date |
|---|---|---|---|---|---|
| Kimi K3 GitHub README | https://github.com/MoonshotAI/Kimi-K3 | official | 2026-09-20 | high | — |
| Kimi API platform — model list | https://platform.kimi.ai/docs/models.md | official | 2026-09-20 | high | — |
| Kimi API — Chat Completions spec | https://platform.kimi.ai/docs/api/chat.md | official | 2026-09-20 | high | — |
| Kimi K3 launch blog | https://www.kimi.com/blog/kimi-k3 | official | 2026-09-20 | high | 2026-07-16 |