Moonshot AI
Kimi K2
Open SourceKimi K2 (0905) is the flagship model from Moonshot AI, a Beijing-based startup — and the current default model on T3.chat. It's a Mixture-of-Experts model with 1 trillion total parameters and 32 billion active per forward pass, released September 2025. Kimi K2 scores an AA Intelligence Index of 31 (#6 of 36 in open-weights large models) and is available as open weights under a permissive license. At $0.39/$1.90 per 1M tokens via the Moonshot API with a 262K context window, it offers strong capability per dollar. Also available in a dedicated Thinking mode for complex reasoning. A newer version (Kimi K2.5) has since launched.
Context window
262K tokens
API (blended)
$0.77/1M
Consumer access
Free (limited)
Multimodal
Text only
Strengths
- +Open weights (1T parameters, 32B active MoE) — can self-host or fine-tune
- +AA Intelligence Index 31 — #6 of 36 in open-weights large model class
- +262K context window — more than double GPT-5.2's 400K is better, but Kimi K2's 262K beats many open models
- +Aggressive pricing: $0.39/$1.90 per 1M — strong value for API pipelines
- +Thinking mode available for complex reasoning tasks
- +T3.chat's default model — widely accessible and tested by many users
Weaknesses
- -Chinese company (Moonshot AI, Beijing) — data jurisdiction concerns for sensitive workloads
- -Newer Kimi K2.5 released — this model is one generation behind as of Feb 2026
- -AA Index 31 is significantly below frontier models (Gemini 3.1 Pro: 57, Claude Opus: 53)
- -Consumer interface (kimi.ai) is primarily Chinese-language — not ideal for English-first users
- -Limited Western enterprise support and SLAs
Best for
Not ideal for
Pricing details
Subscription plans
API pricing
Prices verified February 2026. LLM pricing changes frequently — verify at the provider's site before budgeting.
Last updated: February 2026