Kimi K3 at a glance.
From Moonshot AI's own pages, read Oct 20, 2026, so company-reported. A token is a word or part of a word. Prices can change; check the maker's page before you build on them.
- Made by
- Moonshot AIBased in China
- Released
- Not stated
- What it is for
- “Kimi K3 is an open-weight, native multimodal agentic model and our most capable model to date.”Hugging Face model card moonshotai/Kimi-K3, Oct 20, 2026 · in Moonshot AI's words
- Reads at once
- 1M tokensIts context window: how much text it can take in one go
- Takes in
- text, image, video
- Download
- Yes, open weightsLicense: Kimi K3 License
- Size
- 2.8T parameters, 104B activeHugging Face model card moonshotai/Kimi-K3, Oct 20, 2026
- Run it yourself
- Can I run Kimi K3 on my computer?
$3per million tokens in$15per million tokens out
Input price is cache miss; cache hit $0.30. Cache writes $3.00 (5-minute TTL) or $6.00 (1-hour TTL). Prices exclude taxes. Cheaper siblings: kimi-k2.7-code $0.95/$4.00, kimi-k2.6 $0.95/$4.00.
Kimi API platform: Chat model pricing, Oct 20, 2026 · company-reported
Where it ranks.
Its position on each of the 21 tests the models page uses, by testers other than the maker. Epoch AI's and Vectara's scores are shown under their open licenses; for LMArena, Scale and Vals we give the position and link to their own leaderboard. Each list is a snapshot with its date.
| Test | Rank | Score |
|---|---|---|
| Epoch Capabilities IndexEpoch AI, Oct 20, 2026 | #17 | 157.4range 155.1 to 160.11 |
| Vals IndexVals AI, Oct 19, 2026 | #18 | On Vals AI |
| Test | Rank | Score |
|---|---|---|
| LMArena WebDevLMArena, Oct 19, 2026 | #16 | On LMArena |
| SWE-Bench ProScale AI, Sep 22, 2026 | #4 | On Scale AI |
| Vibe Code BenchVals AI, Oct 18, 2026 | #16 | On Vals AI |
| LMArena coding chatsLMArena, Oct 15, 2026 | #12 | On LMArena |
| Test | Rank | Score |
|---|---|---|
| LMArena textLMArena, Oct 15, 2026 | #18 | On LMArena |
| LMArena creative writingLMArena, Oct 15, 2026 | Not in the top 20 | |
| MultiChallengeScale AI, Feb 3, 2026 | Not in the top 15 |
| Test | Rank | Score |
|---|---|---|
| LMArena Agent ArenaLMArena, Oct 19, 2026 | Not in the top 20 | |
| Remote Labor IndexScale AI | Not in the top 16 | |
| Terminal-Bench 4.0Vals AI, Oct 19, 2026 | Not in the top 20 | |
| CUA-bench (games)Vals AI, Oct 13, 2026 | Not in the top 9 |
| Test | Rank | Score |
|---|---|---|
| LMArena mathLMArena, Oct 15, 2026 | #16 | On LMArena |
| Humanity's Last ExamScale AI | #14 | On Scale AI |
| ProofBenchVals AI, Oct 12, 2026 | #12 | On Vals AI |
| Test | Rank | Score |
|---|---|---|
| LMArena visionLMArena, Oct 19, 2026 | Not in the top 20 | |
| LMArena documentsLMArena, Sep 13, 2026 | Not in the top 20 |
| Test | Rank | Score |
|---|---|---|
| GPQA DiamondVals AI, Sep 1, 2026 | #12 | On Vals AI |
| MMMU ProVals AI, Sep 1, 2026 | #11 | On Vals AI |
| Test | Rank | Score |
|---|---|---|
| Vectara hallucination testVectara, Oct 16, 2026 | Not in our copy of the list |
“Not in the top” means it isn't among the models in our copy of that list, which runs to the position shown; it may not have been tested. Vectara's test ranks by how rarely a model adds facts, so lower is better. Positions can be shared when scores tie. Epoch's range is its margin of uncertainty. Epoch AI and Vectara data: CC BY 4.0 and Apache-2.0.
Our note: Quickstart: 'designed for frontier intelligence scenarios including long-horizon coding, knowledge work, and reasoning.' Release date not stated on pages read; the Hugging Face repo was created 2026-06-13. Card: 'We release the full Kimi K3 model weights under the Kimi K3 License.'