All models › moonshot
Arena score1474#67 of 145 in Coding, paid models included
Size1026.9 B33.3 B active per token (mixture of experts)
4-bit download622 GBweights only, Q4_K_M
Max context256Ktokens, from the model config
Where to use it for free
- The lab's own chat: www.kimi.com (free with daily limits)
- Download and run it at home: official weights · GGUF files for Ollama and LM Studio
Arena scores
| Arena | Score | Rank | Votes |
|---|---|---|---|
| Coding | 1474 | #67 of 145 | 5K |
Rank among all models, paid ones included. Source: Arena leaderboard dataset.
Speed on popular hardware
| Hardware | GPU memory | Fits? | Speed |
|---|---|---|---|
| RTX 3060 12 GB | 12 GB | Does not fit | — |
| RTX 4060 Ti 16 GB | 16 GB | Does not fit | — |
| RTX 3090 | 24 GB | Does not fit | — |
| RTX 4090 | 24 GB | Does not fit | — |
| RTX 5090 | 32 GB | Does not fit | — |
| RX 7900 XTX | 24 GB | Does not fit | — |
| 2 × RTX 3090 | 48 GB | Does not fit | — |
| 4 × RTX 3090 | 96 GB | Does not fit | — |
| Mac M4 Max 128 GB | 96 GB | Does not fit | — |
| Ryzen AI Max+ 395 128 GB | 96 GB | Does not fit | — |
| 8 × H100 SXM | 640 GB | Fits in GPU | 86 tok/s |
Estimate at 4-bit with 32K context, 64 GB of DDR5 for offload. Speed ≈ memory bandwidth × 70% ÷ data read per token; real numbers vary by about ±30%.
GPU memory needed
| Level | 8K context | 32K context | 128K context | Quality vs full |
|---|---|---|---|---|
| Q3_K_M | 515 GB | 516 GB | 523 GB | +4.7% noticeable loss |
| Q4_K_M | 623 GB | 625 GB | 632 GB | +1.2% very small loss |
| Q5_K_M | 731 GB | 733 GB | 740 GB | +0.4% near-lossless |
| Q6_K | 843 GB | 845 GB | 852 GB | +0.2% near-lossless |
| Q8_0 | 1092 GB | 1094 GB | 1101 GB | ≈ 0% same as full |
| F16 | 2055 GB | 2057 GB | 2064 GB | ≈ 0% same as full |
Weights + context cache (from the model's architecture) + runtime. Quality: increase in perplexity against the full model, from llama.cpp tests, scaled by model size. Simulate your own setup.
Coding score over time
1500160017001800
MayJunJulAugSeptOct
Kimi K2.7 Code 1474Best paid model
Arena score from 20 Apr 2026 to 18 Oct 2026.
Similar free models in Coding
- MiMo-V2.5-Pro 1481 · 1023.2 B
- MiniMax-M3 1482 · 427 B
- Hy3 1507 · 298.8 B
- GLM-5.1 1508 · 753.9 B
- MiMo-V2.5 1439 · 310.8 B
- Kimi K2.6 1510 · 1026.9 B