Kimi K2
moonshot/kimi-k2- Moonshot AI
- Family · Kimi K2
- Released 11 Jul 2025
- Open weights
- Text → Text
- Tools
Spec & list price
Context128K
Max output96K
Input / 1M$0.57
Output / 1M$2.304× input
Scores · standing
Intelligence12.7#202/489
Usage
Tokens / day152.9M▼ 14% 7d
Usage rank#279
Requests / day22.5K
Tool calls / req0.09
Per request · 30 days
Avg input4.2K0.3× median
Avg output4700.7× median
Cache hit58%0.8× median
Reasoning share60%about median
Tool-call failures1.4%0.7× median
Consumption
Peak 571.2M 12 OctLow 120.1M 1 NovRank today #279
News & trends
6 mentions this week ▼ 60%Kimi K2 family · videos, posts, articles · 30 days
Trending · 14 days · Kimi K2 family
- Anthropic Haiku 5.5 vs Moonshot Kimi K2.6 draw “a beach scene” #shorts199 views
- Kimi k2.57 upvotes
- Kimi K2利用者は12月末までに移行可否を確認 #shorts281 views
- Stuck between GLM 4.7, 4.6, and Kimi K2.5. Also, which provider should I use between OpenRouter and DeepInfra?49 upvotes
News · latest from Moonshot AI
- ReplaySSM: Rethinking State Storage for Faster Hybrid-Model DecodingNVIDIA · 12 Oct
- Together Link: open models in the harness you already use. Start with one command today.Together AI · 5 Oct
- Kimi K3: the complete developer guideTogether AI · 1 Aug
- Together AI announces strategic partnership with Moonshot AI to natively serve Kimi modelsTogether AI · 29 Jul
Benchmarks
bar = share of the top score · Benchmark data: Artificial AnalysisSpeed · median across providers, latest day
Output speed#272/33440 tok/s1.6× slower than median
First token#2001.31 s1.3× longer than median
#22 cheapest of 29 with a similar score
Value
Intelligence Index
Blended price per 1M tokens (log)
Family progress
Intelligence Index
- Family releases27.0
Technical details
- Tokenizer
- Other
- Knowledge cutoff
- 2024-12-31
- License
- other
- Parameters
- 1T
- Moderation
- none
- Downloads · likes
- 314.6K · 3.0K
Request parameters 11
- tools
- tool_choice
- temperature
- top_p
- top_k
- max_tokens
- stop
- seed
- frequency_penalty
- presence_penalty
+1 more
- repetition_penalty
When to use
Best for
- Math-heavy and knowledge Q&A
- Tool-calling assistants
- Coding help and code explanation
- Self-hosted open-weight deployments
Think twice for
- Structured output guarantees (not listed)
- Image understanding (text-only input)
- Very long documents
- Newer builds where later Kimi versions exist
Research papers
- Can LLMs Extract Architectural Design Decisions from Source Code Commits? - A Preliminary Exploratory Study3 Sep
- MORFES: A Benchmark for Productive Inflectional Competence in Modern Greek30 Jul
- Kimi K3: Open Frontier Intelligence27 Jul
- Decoding the Skew: Distribution-Aware MoE Inference with Adaptive Kernel Dispatch25 Jul
- Reading Between the Dots: Decoding Hidden Computation across Filler Tokens3 Jul
- Not All Refusals Are Equal: How Safety Alignment Fails Cybersecurity at Scale2 Jul
- SenFlow: Inter-Sentence Flow Modeling for AI-Generated Text Detection in Hybrid Documents17 Jun
- To Select or not to Select, that is the Question: Distilling Robot Skill Prediction into a Small Ensemble20 May
Social
Most watched and discussed · 90 days
- Kimi K2利用者は12月末までに移行可否を確認 #shorts60秒AI検索ニュース · 281 views · 23 Oct
- Kimi K2 實測：AI 只在有人看的時候才乖？企業 AI 落地研究所 · 9 views · 26 Oct
- #11 Windsurf Acquired, Kimi K2, Kiro, Canva MCP, ChatGPT Agents, AI GirlfriendCmd+AI · 7 views · 20 Oct
- Muon Optimizer: The Quiet Breakthrough Behind Kimi K2 and DeepSeek-V4The AI Signal · 1 views · 29 Oct
Same family
Kimi K2- 22.0
- 27.0
- 23.5
- 15.3
Alternatives
other vendors · similar capability- 13.1
- 15.6
- 10.2
- 13.7
Using Kimi K2 at volume?
Sign in to ask for a volume discount. You will continue in the OSMEXO dashboard.