Kimi K2.6
- In Input modalities: TextImageVideo Out Output modalities: Text
- Reasoning
- Reasoning can be off
- Tool use
- Structured output
- Open weights
Prices
6 offerings · prices as listed| Conditions/ 1M tokens | Input | Cache read | Output |
|---|---|---|---|
| Standard | $0.95 | $0.16 | $4.00 |
Prices from Moonshot AI platform pricing (opens in a new tab) (checked Nov 8, 2026)
| Conditions/ 1M tokens | Input | Cache read | Output |
|---|---|---|---|
| Standard | ¥6.50 | ¥1.10 | ¥27.00 |
Prices from Kimi API 开放平台：模型推理价格 (opens in a new tab) (checked Nov 8, 2026)
| Conditions/ 1M tokens | Input | Output |
|---|---|---|
| Standard | ¥6.50 | ¥27.00 |
Prices from 阿里云百炼：模型价格 (opens in a new tab) (checked Nov 9, 2026)
kimi/kimi-k2.6| Conditions/ 1M tokens | Input | Output |
|---|---|---|
| Standard | ¥6.50 | ¥27.00 |
Prices from 阿里云百炼：模型价格 (opens in a new tab) (checked Nov 9, 2026)
Standard: the price a request with no conditions pays (the headline price elsewhere on the site).
Details
- Knowledge cutoff
- Jan 2025
- Licence
- Modified MIT License
- Family
- kimi-k
- Model ID
Not stated on Moonshot AI’s pages: max input, reasoning efforts, max thinking budget, built-in tools, aliases.
Run locally
Weights: moonshotai/Kimi-K2.6- Parameters 1.03T
- Active 32B · 384 experts, 8 per token
- Native precision BF16
- Layers 61
- KV cache per token 69 KB
- Config positions 256K (config.json max_position_embeddings)
- API context 256K (maker’s API)
- Needs
- ≈ 534.0 GB (weights 508.1 + KV cache 0.5 + overhead 25.4)
- Usable
- 21.6 GB (90% of its memory)
- Try
- it’s too large for this hardware; see the smallest that runs it below
| Precision | Weights | Needs at 8K | On your hardware | Longest context |
|---|---|---|---|---|
| BF16 (native)maker’s file | 554.3 GB | ≈ 582.6 | ✕Doesn’t fit | — |
| Q8_0 | 1016.1 GB | ≈ 1067.5 | ✕Doesn’t fit | — |
| FP8 | 956.4 GB | ≈ 1004.7 | ✕Doesn’t fit | — |
| Q6_K | 785.4 GB | ≈ 825.2 | ✕Doesn’t fit | — |
| Q5_K_M | 680.2 GB | ≈ 714.8 | ✕Doesn’t fit | — |
| Q5_0 | 657.5 GB | ≈ 690.9 | ✕Doesn’t fit | — |
| Q4_K_M | 581.0 GB | ≈ 610.6 | ✕Doesn’t fit | — |
| Q4_0 | 538.0 GB | ≈ 565.4 | ✕Doesn’t fit | — |
| IQ4_XS | 508.1 GB | ≈ 534.0 | ✕Doesn’t fit | — |
| Q3_K_Mquality drops | 467.4 GB | ≈ 491.3 | ✕Doesn’t fit | — |
| IQ3_XXSquality drops | 365.8 GB | ≈ 384.6 | ✕Doesn’t fit | — |
| Q2_Kquality drops | 358.6 GB | ≈ 377.1 | ✕Doesn’t fit | — |
| IQ2_XXSquality drops | 246.3 GB | ≈ 259.1 | ✕Doesn’t fit | — |
Click a row to choose a precision. “Maker’s file” sizes are the maker’s own files; others are parameters × bits per weight.
Compared with other hardware
Each at the highest precision it runs. Click one for details.
Smallest that runs it (Q4_K_M, 8K): no single listed machine; about 9 × 80 GB GPUs.
Run commands
The maker publishes no files at this precision. Choose its native precision or one of its own builds, or quantise it yourself with llama.cpp.
Memory to fine-tune
Batch 1, 2,048-token sequences, gradient checkpointing; LoRA and QLoRA train rank-16 adapters on the attention projections.
Estimates; facts from the maker’s Hugging Face repositories, as of 2026-11-11. How it’s estimated
Cited scores
From third-party leaderboards, not measured by us- Epoch Capabilities Index 151.6
#54 of 118 listed models · CI 149.8–153.4
- LMArena Text 1,461
#40 of 163 listed models · CI 1,456–1,465 · 39,576 votes
- Epoch AI · Chess Puzzles 20%
#66 of 85 listed models · ±4.02
- Epoch AI · EBR-bench 2.4%
#27 of 28 listed models
- Epoch AI, 'Capabilities & benchmarking'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/benchmarks'. Licensed CC BY 4.0. · License · as of Oct 7 – Nov 5, 2026
- LMArena (Arena), leaderboard-dataset on Hugging Face, licensed CC BY 4.0 · License · as of Nov 8, 2026
- EQ-Bench by Samuel J. Paech (eqbench.com; EQ-bench/EQ-bench-site on GitHub), MIT License. · License · as of Oct 24 – 28, 2026
- changes we made
Provider-reported scores
Published by Moonshot AI, under the conditions it states| Benchmark | Setting | Score |
|---|---|---|
| Humanity's Last Exam | full set, no tools, thinking | 34.7% |
| full set, with search/code/browsing tools, thinking | 54% | |
| BrowseComp | thinking, with tools, discard-all context management | 83.2% |
| Agent Swarm mode | 86.3% | |
| OSWorld-Verified | thinking | 73.1% |
| Terminal-Bench 2.0 | Terminus-2 harness, preserve thinking | 66.7% |
| SWE-bench Pro | in-house SWE-agent-based harness, avg of 10 runs | 58.6% |
| SWE-bench Verified | in-house SWE-agent-based harness, avg of 10 runs | 80.2% |
| LiveCodeBench | v6, thinking | 89.6% |
| AIME 2026 | thinking | 96.4% |
| GPQA Diamond | thinking | 90.5% |
| MMMU-Pro | no tools, thinking, avg@3 | 79.4% |
| with Python tool, thinking, avg@3 | 80.1% |
Changes
All changes →- Available → Deprecated
- Now priced on Alibaba Cloud Model Studio, China Input ¥6.50 · Output ¥27.00
- Now priced on Alibaba Cloud Model Studio, China Input ¥6.50 · Output ¥27.00
- Now priced on Cloudflare Workers AI Input $0.95 · Cache read $0.16 · Output $4.00
Sources
- moonshotai/Kimi-K2.6 model card (opens in a new tab) huggingface.co · checked Nov 6, 2026
- Moonshot AI platform pricing (opens in a new tab) platform.moonshot.ai · checked Nov 8, 2026
- Kimi API 开放平台：模型推理价格in Chinese (opens in a new tab) platform.moonshot.cn · checked Nov 8, 2026
- 硅基流动：模型价格in Chinese (opens in a new tab) siliconflow.cn · checked Nov 8, 2026
- Kimi API introduction (opens in a new tab) platform.kimi.ai · checked Oct 18, 2026
- Kimi API models overview (opens in a new tab) platform.kimi.ai · checked Nov 8, 2026
- Kimi API structured output (opens in a new tab) platform.kimi.ai · checked Nov 8, 2026
- Kimi API responses reference (opens in a new tab) platform.kimi.ai · checked Nov 8, 2026
- Kimi platform changelog (opens in a new tab) platform.kimi.ai · checked Nov 9, 2026
- models.dev: kimi-k2.6 (community data, MIT) (opens in a new tab) github.com · checked Sep 30, 2026
- Kimi K2.6 - Kimi API Platform (opens in a new tab) platform.kimi.ai · checked Nov 8, 2026
- moonshotai/Kimi-K2.6 · Hugging Face (opens in a new tab) huggingface.co · checked Nov 7, 2026
- Cloudflare Workers AI pricing (opens in a new tab) developers.cloudflare.com · checked Nov 7, 2026
- 阿里云百炼：模型价格in Chinese (opens in a new tab) help.aliyun.com · checked Nov 9, 2026
- 模型上下架与更新-大模型服务平台百炼-阿里云_大模型服务平台百炼(Model Studio)-大模型服务平台百炼(Model Studio)-阿里云帮助中心in Chinese (opens in a new tab) help.aliyun.com · checked Nov 4, 2026
- moonshotai/Kimi-K2.6 model card (opens in a new tab) huggingface.co · checked Nov 5, 2026
- moonshotai/Kimi-K2.6 config.json (opens in a new tab) huggingface.co · checked Nov 8, 2026
- moonshotai/Kimi-K2.6 files (opens in a new tab) huggingface.co · checked Nov 6, 2026
- moonshotai/Kimi-K2.6 README.md (opens in a new tab) huggingface.co · checked Nov 6, 2026
Moonshot AI’s official policies: data use, privacy, terms →
Some values on this page are still waiting to be checked against official pages: a filled dot is verified, a hollow one is not yet (with or without a source on file). How we verify data
Moonshot AI · China · moonshot.ai
Related models
Snapshots, the same family and other models from this provider