MMoonshot AI · Released Jul 16, 2026
Kimi K3
Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at...
- Top 9%SciCode
- Top 10%WebDev Arena
- Top 15%ProofBench
- Top 16%GPQA Diamond
- Released
- Jul 16 2026
- Context window
- 1M 944K out
- Input / 1M tokens
- $0.24 from Free
- Output / 1M tokens
- $11.94 from Free
Where to run it
93 providers compared
Free to try on SenseNova (China), ZenMux, Nvidia — free endpoints are often limited-time and rate-limited. Included in SCNet Token Plan, Kimi For Coding (kimi.ai), Umans AI Coding Plan, Volcengine Ark Coding Plan, Kimi For Coding (kimi.com) +1 more. Cheapest: Sail Research at $1.20 / $5.50 per 1M input / output tokens — 81% less than the priciest option.
| Provider | Input | Output | Cache read | Context | Max output | Uptime | vs. best |
|---|---|---|---|---|---|---|---|
| SenseNova (China)free · may be limited-time | Free | Free | Free | 1M | 66K | — | Free |
| ZenMuxfree · may be limited-time | Free | Free | Free | 1M | 131K | — | Free |
| Nvidiafree · may be limited-time | Free | Free | — | 1M | 131K | — | Free |
| SCNet Token Plansubscription plan | Plan | Plan | Plan | 1M | 131K | — | Plan |
| Kimi For Coding (kimi.ai)subscription plan | Plan | Plan | Plan | 262K | 131K | — | Plan |
| Umans AI Coding Plansubscription plan | Plan | Plan | Plan | 1M | 131K | — | Plan |
| Volcengine Ark Coding Plansubscription plan | Plan | Plan | Plan | 1M | 131K | — | Plan |
| Kimi For Coding (kimi.com)subscription plan | Plan | Plan | Plan | 262K | 131K | — | Plan |
| OpenCode Gosubscription plan | Plan | Plan | Plan | 1M | 131K | — | Plan |
| Sail Researchus · FP4 · no tool calling | $1.20 | $5.50 | $0.20 | 1M | 944K | — | Cheapest |
| Sail ResearchFP4 | $1.20 | $5.50 | $0.20 | 1M | 944K | — | Cheapest |
| Kenari | $0.86 | $8.57 | $0.11 | 1M | 131K | — | +22% |
| InferenceNetus · FP4 | $0.24 | $11.94 | $0.21 | 1M | 944K | 100%24h | +39% |
| InferenceNetFP4 | $0.24 | $11.94 | $0.21 | 1M | 944K | 100%24h | +39% |
| RelaceFP4 | $0.30 | $12.00 | $0.25 | 1M | 944K | 99.5%24h | +42% |
| Waferus · MXFP4 | $0.36 | $11.90 | $0.36 | 1M | 944K | 100%24h | +43% |
| Decartmxfp4 · MXFP4 | $1.65 | $8.25 | $0.17 | 1M | 944K | 99.8%24h | +45% |
| Chutesmxfp4 · MXFP4 · no tool calling | $1.85 | $9.50 | $0.18 | 1M | 66K | 95.7%24h | +65% |
| engy | $1.95 | $9.75 | $0.20 | 1.11M | 66K | — | +71% |
| Neuralwatt | $1.95 | $9.75 | $0.20 | 1.05M | 1.05M | — | +71% |
| Fireworksspot | $2.10 | $10.50 | $0.21 | 1M | 944K | 70.7%24h | +85% |
| ainetcafe | $2.10 | $10.50 | $0.30 | 262K | 33K | — | +85% |
| MakoraFP4 | $0.65 | $15.00 | $0.23 | 1M | 944K | 99.4%24h | +86% |
| MorphFP8 | $0.89 | $14.90 | $0.43 | 1M | 944K | 98.9%24h | +93% |
| AkashMLFP4 | $1.20 | $14.00 | $1.20 | 1M | 944K | 99.7%24h | +93% |
| Boundless | $2.30 | $11.40 | $0.23 | 1M | 131K | — | +101% |
| IteraCompute | $2.35 | $11.76 | $0.23 | 1M | 1000K | — | +107% |
| Vancine | $2.40 | $12.00 | $0.24 | 1M | 131K | — | +111% |
| InferenceNetfast · FP4 | $2.04 | $13.20 | $0.35 | 1M | 944K | 99.4%24h | +112% |
| Phala | $2.55 | $12.75 | $0.26 | 1M | 944K | 99.9%24h | +124% |
| DigitalOcean | $2.55 | $12.95 | $0.26 | 1M | 944K | 100%24h | +126% |
| InferenceNetus/fast · FP4 | $2.24 | $14.52 | $0.35 | 1M | 944K | 99.7%24h | +133% |
| ai& | $3.00 | $12.50 | $0.50 | 1M | 131K | — | +136% |
| SiliconFlow | $2.70 | $13.50 | $0.27 | 1M | 262K | — | +137% |
| Wallaby | $2.70 | $13.50 | $0.27 | 1M | 131K | — | +137% |
| Alibaba (China) | $2.83 | $14.13 | $0.28 | 1M | 1M | — | +149% |
| Merge Gateway | $2.90 | $14.00 | $0.30 | 1M | 1M | — | +149% |
| DeepInframxfp4 · MXFP4 | $2.85 | $14.25 | $0.28 | 1M | 944K | 99.9%24h | +151% |
| Moonshot AI (China) | $2.99 | $14.94 | $0.30 | 1M | 1M | — | +163% |
| Cortecs | $3.00 | $15.00 | — | 1M | 1M | — | +164% |
| Modalmxfp4 · MXFP4 | $3.00 | $15.00 | $0.30 | 1M | 944K | 100%24h | +164% |
| Together | $3.00 | $15.00 | $0.30 | 1M | 944K | 100%24h | +164% |
| Moonshot AImxfp4 · MXFP4 | $3.00 | $15.00 | $0.30 | 1M | 944K | 100%24h | +164% |
| BaseTenFP8 | $3.00 | $15.00 | $0.30 | 1M | 262K | 99.6%24h | +164% |
| Fireworks | $3.00 | $15.00 | $0.30 | 1M | 944K | 98%24h | +164% |
| ParasailFP4 | $3.00 | $15.00 | $0.30 | 1M | 944K | 96.9%24h | +164% |
| Amazon Bedrockus-east-2 | $3.00 | $15.00 | $0.30 | 1M | 128K | 94.3%24h | +164% |
| NanoGPT | $3.00 | $15.00 | $1.50 | 1M | 66K | — | +164% |
| Cloudflare AI Gateway | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Moark | $3.00 | $15.00 | — | 1M | 131K | — | +164% |
| Kilo Gateway | $3.00 | $15.00 | $0.30 | 1M | 944K | — | +164% |
| LLM Gateway | $3.00 | $15.00 | $0.30 | 1M | 1M | — | +164% |
| TokenGo | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Umans AI | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Nebius Token Factory | $3.00 | $15.00 | $3.00 | 1.02M | 8K | — | +164% |
| Arcee | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Together AI | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Requesty | $3.00 | $15.00 | $0.45 | 1M | 262K | — | +164% |
| AIHubMix | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| 302.AI | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Jalapeno Cloud | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Fireworks AI | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| ZenMux | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Synthetic | $3.00 | $15.00 | $0.45 | 524K | 66K | — | +164% |
| GitHub Copilot | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Novita AI | $3.00 | $15.00 | $0.30 | 1M | 1M | — | +164% |
| ClinePass | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Perplexity Agent | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Neon | $3.00 | $15.00 | $0.30 | 1M | 66K | — | +164% |
| Vivgrid | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Abacus | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| CrossModel | $3.00 | $15.00 | $0.30 | 1M | 1M | — | +164% |
| Tempr Gateway | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Hugging Face | $3.00 | $15.00 | — | 1M | 131K | — | +164% |
| Ollama Cloud | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| CommandCode | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Ofox | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Impossibl | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| OrcaRouter | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| OpenCode | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| TensorX | $3.00 | $15.00 | $0.75 | 1M | 131K | — | +164% |
| EmpirioLabs AI | $3.00 | $15.00 | $3.00 | 1M | 131K | — | +164% |
| Eden AI | $3.00 | $15.00 | $0.30 | 1M | 131K | — | +164% |
| Opper | $3.00 | $15.00 | — | 1M | 131K | — | +164% |
| Standard Compute | $3.00 | $15.00 | $0.30 | 1M | 50K | — | +164% |
| WaferMXFP4 | $0.24 | $24.00 | $0.23 | 1M | 944K | 100%24h | +172% |
| Melious | $3.20 | $15.98 | $0.84 | 1M | 131K | — | +181% |
| Charm Hyper | $3.27 | $16.33 | $0.33 | 1M | 16K | — | +187% |
| NEAR AI Cloud | $3.30 | $16.50 | $0.33 | 1M | 131K | — | +190% |
| Berget.AI | $3.37 | $16.82 | $1.01 | 328K | 33K | — | +196% |
| Alibaba | $3.45 | $17.25 | $0.34 | 1M | 131K | 98.7%24h | +203% |
| Prime Inference | $3.45 | $17.25 | — | 1M | 944K | — | +203% |
| GreenPT | $3.76 | $18.81 | $0.94 | 1M | 131K | — | +231% |
| AI/ML API | $3.90 | $19.50 | $0.39 | 1M | 131K | — | +243% |
| Tinfoil | $4.00 | $20.00 | $0.80 | 262K | 131K | — | +252% |
| Fireworksfast | $4.50 | $22.50 | $0.45 | 1M | 944K | 98.9%24h | +296% |
| Fireworksus | $4.50 | $22.50 | $0.45 | 1M | 944K | 93%24h | +296% |
| Venice AI | $4.50 | $22.50 | $0.45 | 1M | 131K | — | +296% |
| Vercel AI Gateway | $4.50 | $22.50 | $0.45 | 1M | 131K | — | +296% |
| Pioneer | $4.50 | $22.50 | $0.45 | 1M | 131K | — | +296% |
| DevPass (LLM Gateway) | $4.50 | $22.50 | $0.45 | 1.04M | 131K | — | +296% |
| Inco | $6.00 | $30.00 | $0.60 | 1M | 131K | — | +427% |
Value
Price vs. performance
Kimi K3 Other Moonshot AI Other labs Value frontier · GPQA Diamond vs. input price per 1M tokens (log scale) · Drag to pan, pinch or Ctrl + scroll to zoom, double-click to zoom in.
Benchmarks
25 benchmark scores
| Benchmark | Score | Rank | Leader |
|---|---|---|---|
| FrontierSWESoftware engineering · Ultra-long-horizon engineering tasks with a 20-hour budget each | 25.9% | #12of 19 | GPT-6 Astra65.5% |
| DeepSWESoftware engineering · Original long-horizon tasks across active open-source repos | 68.5% | #8of 26 | GPT-6 Astra74.1% |
| FrontierCodeSoftware engineering · Mergeable fixes for hard open-source issues | 44.2% | #15of 41 | Claude Opus 5.554.6% |
| GBAEvalSoftware engineering · Build a Game Boy Advance emulator from scratch | 48.4% | #9of 23 | Claude Opus 579.6% |
| SciCodeSoftware engineering · Research-coding problems from the natural sciences | 59.5% | #9of 102 | Claude Opus 5.566.9% |
| WeirdML v3Software engineering · Unfamiliar ML and data tasks with vague goals | 7.5% | #14of 22 | GPT-6 Astra42.2% |
| APEX-AgentsAgents & computer use · Long professional tasks in banking, consulting and law | 47.8% | #28of 51 | Gemini 4 Argon82.4% |
| Vending-Bench 2Agents & computer use · Run a simulated vending business for a year (final balance) | $5,165 | #27of 61 | GPT-6 Astra$15,515 |
| PostTrainBenchAgents & computer use · Post-train small models under a fixed compute budget | 32.0% | #5of 11 | Claude Fable 541.8% |
| GDP.pdfWriting & content · Answer questions about real professional PDFs | 19.0% | #24of 38 | GPT-6.1 Sol35.0% |
| SimpleQA VerifiedWriting & content · Factual accuracy on 1,000 short questions | 50.6% | #25of 79 | Gemini 4 Argon76.0% |
| WebDev ArenaDesign & visual · People vote on which model builds the better website (Elo) | 1674 | #9of 90 | Claude Opus 5.51820 |
| Blueprint-Bench 2Design & visual · Turn apartment photos into 2D floor plans | 29.5% | #19of 32 | Gemini 4 Argon54.4% |
| Furniture AssemblyDesign & visual · Spot mistakes in photos of IKEA builds | 34.2% | #20of 34 | Claude Opus 5.583.3% |
| GPQA DiamondReasoning & knowledge · PhD-level biology, chemistry and physics questions | 93.1% | #19of 122 | Gemini 4 Argon96.4% |
| ARC-AGI-2Reasoning & knowledge · Few-shot abstract reasoning on grid puzzles | 60.4% | #33of 75 | GPT-6 Astra95.0% |
| ARC-AGI-1Reasoning & knowledge · Pattern induction from input–output examples | 94.5% | #16of 75 | Gemini 4 Argon98.5% |
| SimpleBenchReasoning & knowledge · Common-sense trick questions about space, time and people | 60.7% | #21of 58 | Claude Fable 581.9% |
| CritPtReasoning & knowledge · Open-ended, research-level physics problems | 23.4% | #18of 111 | GPT-5.6 Sol32.3% |
| Chess PuzzlesReasoning & knowledge · Find the single best move in 100 new puzzles | 37.0% | #38of 83 | GPT-6 Astra76.0% |
| LMCAReasoning & knowledge · Judge the quality of conceptual arguments like experts | 52.7% | #20of 123 | Claude Sonnet 5.568.3% |
| FrontierMathMaths · Expert-written problems, undergrad to early research (tiers 1–3) | 72.2% | #23of 87 | GPT-6.1 Sol93.7% |
| FrontierMath Tier 4Maths · Exceptionally hard research-level problems | 39.0% | #25of 64 | GPT-6.1 Sol100.0% |
| ProofBenchMaths · Graduate maths proofs verified in Lean 4 | 87.0% | #10of 70 | Claude Sonnet 5.5100.0% |
| OTIS Mock AIMEMaths · Competition-style problems harder than MATH Level 5 | 97.2% | #29of 120 | Gemini 4 Argon100.0% |
Specs
Details
- Model ID
moonshotai/kimi-k3- Developer
- Moonshot AI
- Release date
- Jul 16, 2026 ·
- Context window
- 1,048,576 tokens
- Max output
- 943,718 tokens
- API price
- $0.24 / $11.94 per 1M input / output tokens
- Providers
- 93
- Input types
- text, image, video
- Reasoning
- Yes
- Tool calling
- Yes
- Open weights
- Yes
- Free tier
- No
Release date and prices from OpenRouter's catalog; providers from OpenRouter and models.dev. Buying directly from a provider can cost a little differently. Speed, latency and uptime are measured on OpenRouter only; uptime is the share of OpenRouter requests each provider served over the last day. Benchmark scores are the best result per benchmark from Epoch AI, licensed CC BY 4.0. Rank counts every model tracked here with a score.