Kimi K2.6
Kimi K2.6Alternatives?
CheaperMiniMax M3thinkingOpen+2%-37%$0.51No strict JSON+1
BetterGPT-6 Astralow+2%same$0.82Monitoring can return HTTP 403+2
OpenQwen3.8 2.4T A95Bthinking+3%+169%$2.16Memory (4-bit): 616 GB → 1,468 GB+6
CheaperDeepSeek V4 Pro 0813maxOpen+2%-16%$0.67Memory (4-bit): 616 GB → 990 GB+6CheaperGPT-5.6 Solmedium+2%-37%$0.50MiniMax M3thinkingOpen+2%-37%$0.51No strict JSON+1DeepSeek V4 Pro 0813maxOpen+2%-16%$0.67Memory (4-bit): 616 GB → 990 GB+6GPT-5.6 Solmedium+2%-37%$0.50Qwen3.8 FlashthinkingOpen+1%-53%$0.37max_tokens range · QwenCloud+5
Gemini 3.8 Flashlowsame-12%*$0.75 / $3.75Sampling parameters are ignored+1GPT-6 Astralow+2%same$0.82Monitoring can return HTTP 403+2
Grok 4.5high+2%+10%$0.88Gemini 3.8 Flashmedium+3%+16%*$0.93Sampling parameters are ignored+1Grok 4.6medium+3%+43%$1.14GPT-5.6 Solxhigh+2%+48%$1.18Qwen3.8 2.4T A95BthinkingOpen+3%+169%$2.16Memory (4-bit): 616 GB → 1,468 GB+6MiniMax M3thinkingOpen+2%-37%$0.51No strict JSON+1DeepSeek V4 Pro 0813maxOpen+2%-16%$0.67Memory (4-bit): 616 GB → 990 GB+6Qwen3.8 FlashthinkingOpen+1%-53%$0.37max_tokens range · QwenCloud+5
GLM 5.3maxOpensame+151%$2.01No image input+4Kimi K3maxOpen+3%+150%$2.00Memory (4-bit): 616 GB → 1,668 GB+4Kimi K2.7 CodethinkingOpen-2%-32%$0.54Instant mode is not supported · Self-hosted / vLLM / SGLangKimi K2.5thinkingOpen-4%—$0.60 / $3Kimi K2 ThinkingthinkingOpen-8%—$0.47 / $2No image inputKimi K2 0905non-thinkingOpen-16%—$0.40 / $2No image inputKimi K2 0711non-thinkingOpen-16%—$0.50 / $2Context: 256K → 128K+2Kimi K3.1 Pron/a—$2 / $10Kimi K3.1 Pro Highspeedn/a—$4 / $20All Moonshot AI models →Run it yourself
estimateHistory
5 versionsBenchmarks
in Chat & KnowledgeFAQ
How much does Kimi K2.6 cost?›
$0.95 per 1M input tokens and $4 per 1M output tokens; cached input $0.16 per 1M; cheapest via Chutes at $0.45 / $1.95. A typical benchmark task costs about $0.80 at default effort.
Is Kimi K2.6 free to use?›
There is no free API; the cheapest paid price is $0.45 per 1M input tokens. The weights are open, so you can run it yourself.
How good is Kimi K2.6, and what is it best at?›
#55 of 334 in Chat & Knowledge, #63 of 132 in Coding, #73 of 133 in Agents & Tool Use, #44 of 195 in Writing & Language, #23 of 124 in Vision & Documents. Strongest at: Tool calling, Reasoning, Science, Long context, Knowledge, Instructions, Vision.
Is Kimi K2.6 better than Kimi K2.5?›
On GPQA Diamond, Kimi K2.6 scores 91.1 vs 87.9 — higher.
How can I use Kimi K2.6?›
Through the API of AtlasCloud, Azure AI Foundry, Chutes, Cloudflare, CoreWeave and 10 more.
What is the context window of Kimi K2.6?›
256K tokens, with up to 236K output tokens.
When was Kimi K2.6 released? Is there a newer version?›
Released 2026-04-20. It has been superseded by Kimi K3.
Is Kimi K2.6 open source?›
Yes, the weights are open (modified-mit licence).
Can I run Kimi K2.6 locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Kimi K2.6 read images?›
It accepts text and images as input.
What is Kimi K2.6?›
Kimi K2.6 is an open-weights language model from Moonshot AI, released Apr 2026, the successor to Kimi K2.5.
What are the best alternatives to Kimi K2.6?›
MiniMax M3 (cheaper), GPT-6 Astra (better), Qwen3.8 2.4T A95B (open weights).