kimi-k2-thinking
by Poe · family kimi-thinking · listed Nov 2025RT
Kimi reasoning model for long-horizon research, planning, and tool use
Current rates USD per 1M tokens
Input—
Output—
Cache read—
Cache write—
Against the market
On the record
- Model id
novita/kimi-k2-thinking- Context window
- 256K tokens
- Max output
- —
- Input modalities
- text
- Output modalities
- text
- Reasoning
- yes
- Tool calls
- yes
- Open weights
- no
- Knowledge cutoff
- —
- Released
- 2025-11-07
- Last updated
- 2025-11-07
Also listed at
- Kimi For Codingfree in · free out
- NanoGPT$0.30 in · $1.20 out
- OpenCode Zen$0.40 in · $2.50 out
- Vercel AI Gateway$0.47 in · $2.00 out
- Kilo Gateway$0.47 in · $2.00 out
- Helicone$0.48 in · $2.00 out
- IO.NET$0.55 in · $2.25 out
- Alibaba (China)$0.574 in · $2.29 out
- LLM Gateway$0.574 in · $2.29 out
- 302.AI$0.575 in · $2.30 out
- Hugging Face$0.60 in · $2.50 out
- NovitaAI$0.60 in · $2.50 out
- Moonshot AI$0.60 in · $2.50 out
- ZenMux$0.60 in · $2.50 out
- Merge Gateway$0.60 in · $2.50 out