Kimi K2.6
Kimi K2.6 is an open-weights language model from Moonshot AI, released April 20, 2026. It has 1T parameters (mixture-of-experts, 33B active). The context window is 256K tokens. On the Epoch Capabilities Index it scores 151.6, rank 62 of 256 rated models. API price: $0.95 input / $4 output per million tokens.
Check it against your own PC →
Can you run Kimi K2.6 locally?
Even with 32 GB of VRAM and 32 GB of RAM it is too large to run at a usable speed on a single consumer GPU. Estimates assume 32 GB of DDR5 RAM, a 16K context and llama.cpp-style GGUF quantization. Enter your exact hardware in the live app.
| GPU memory | Verdict | Best quant | VRAM | RAM | tok/s |
|---|---|---|---|---|---|
| 6 GB VRAM | Too large | – | – | – | – |
| 8 GB VRAM | Too large | – | – | – | – |
| 12 GB VRAM | Too large | – | – | – | – |
| 16 GB VRAM | Too large | – | – | – | – |
| 24 GB VRAM | Too large | – | – | – | – |
| 32 GB VRAM | Too large | – | – | – | – |
Ready-made quantizations on Hugging Face: unsloth/Kimi-K2.6-GGUF · bartowski/moonshotai_Kimi-K2.6-GGUF · DevQuasar/moonshotai.Kimi-K2.6-GGUF · AesSedai/Kimi-K2.6-GGUF · baa-ai/Kimi-K2.6-RAM-344GB-GGUF · baa-ai/Kimi-K2.6-RAM-GGUF
Strengths by category
ModelTape's 0–100 fit on Epoch AI benchmark results.
| Category | Score | Based on |
|---|---|---|
| Reasoning | 12.8 | Chess Puzzles, Mystery Game Puzzles |
| Math | 42.9 | ProofBench, OTIS Mock AIME 2024-2025, FrontierMath-Tiers-1-3-v2-Private, FrontierMath-Tier-4-v2-Private |
| Knowledge | 71.6 | GPQA diamond, SimpleQA Verified |
Benchmark results
| Benchmark (Epoch AI) | Score |
|---|---|
| OTIS Mock AIME | 96.1% |
| GPQA Diamond | 91.4% |
| SWE-bench Verified | 76.7% |
| FrontierMath (Tiers 1–3) | 60.2% |
| SimpleQA Verified | 34.9% |
| FrontierMath-Tier-4-v2-Private | 25.6% |
| ProofBench | 16.0% |
| Chess Puzzles | 15.8% |
| Mystery Game Puzzles | 9.7% |
| OSWorld 2.0 | 4.6% |
| EBR-bench | 2.4% |
| Furniture Assembly | 0.0% |
Links
Announcement · Hugging Face: moonshotai/Kimi-K2.6 · Epoch AI benchmarks
Models with a similar score
Muse Spark Open · Gemini 3.5 Flash · Gemini 3.6 Flash · Muse Spark 1.2 · Grok 4.5 · Muse Spark 1.1 · Mistral Large 4 · Qwen3.7-Max
More from Moonshot AI
Kimi K3.1 Pro · Kimi K3 · Kimi K2.7 Code · Kimi K2.5 · Kimi K2 Thinking · Kimi Linear · Kimi K2 (Jul 2025) · Kimi-VL