{"schema":"postcutoff/model@1","as_of":"2026-11-13T13:49:00+01:00","url":"https://postcutoff.com/m/kimi-k2-thinking/","md":"https://postcutoff.com/m/kimi-k2-thinking/index.md","disclosure":{"written_by":"AI agents (Claude Opus 5.5 in Claude Code)","editor":"human editor","policy":"https://postcutoff.com/about/"},"license":null,"id":"kimi-k2-thinking","name":"Kimi K2 Thinking","org":"Moonshot AI","family":"Kimi K2","released":"2025-11","status":"retired","type":"reasoning-llm","modality_in":["text"],"modality_out":["text"],"open_weights":true,"model_license":"modified-mit","context_window":262144,"max_output":null,"knowledge_cutoff":null,"pricing":null,"price_line":"Price not published","access":[{"provider":"Kimi API (discontinued)","model_id":"kimi-k2-thinking / kimi-k2-thinking-turbo","endpoint":"https://api.moonshot.ai/v1","docs":"https://platform.kimi.ai/docs/models"},{"provider":"OpenRouter","model_id":"moonshotai/kimi-k2-thinking","url":"https://openrouter.ai/moonshotai/kimi-k2-thinking"},{"provider":"Hugging Face","url":"https://huggingface.co/moonshotai/Kimi-K2-Thinking"}],"capabilities":[{"name":"Long tool-call chains","detail":"Interleaves reasoning with function calls and stays coherent across 200-300 sequential tool calls (vs 30-50 for earlier models, per Moonshot).","first":false,"discovered":"launch","source":"https://huggingface.co/moonshotai/Kimi-K2-Thinking"},{"name":"Native INT4 via quantization-aware training","detail":"QAT in post-training gives a lossless ~2x speed-up at INT4 on a 1T/32B-active MoE.","first":false,"discovered":"launch","source":"https://huggingface.co/moonshotai/Kimi-K2-Thinking"},{"name":"Heavy mode","detail":"Parallel 8-trajectory rollout with reflective aggregation used for top benchmark results (HLE, BrowseComp).","first":false,"discovered":"launch","source":"https://huggingface.co/moonshotai/Kimi-K2-Thinking"}],"entry":null,"notes":"kimi-k2 series (incl. K2 Thinking, K2-0905, K2-0711) discontinued on the Kimi API on 2026-05-25; Moonshot recommends kimi-k3. Still available as open weights and via third parties. Pricing not verified.","verified":"2026-09-29","body_md":"Landmark open-weight thinking agent (Nov 2025). Use for self-hosting or research; for API use kimi-k3 or kimi-k2.6.\n\nSources: https://huggingface.co/moonshotai/Kimi-K2-Thinking · https://platform.kimi.ai/docs/models","page_url":"https://postcutoff.com/m/kimi-k2-thinking/","events_after":null,"major_after":null,"historic_after":null,"missing_at_launch":null,"events_since_release":1970,"you_url":null,"briefings":null}