Kimi K2.5 Thinking
by NanoGPT · family kimi-k2 · listed Jan 2026RTVW
Earlier Kimi frontier model for long-context agents, coding, and multimodal work
Current rates USD per 1M tokens
Input$0.30
Output$1.90
Cache read$0.15
Cache write—
Against the market
Input price vs. maker flagships ($/1M tokens)Output price vs. maker flagships ($/1M tokens)On the record
- Model id
moonshotai/kimi-k2.5:thinking- Context window
- 256K tokens
- Max output
- 66K tokens
- Input modalities
- text, image
- Output modalities
- text
- Reasoning
- yes
- Tool calls
- yes
- Open weights
- yes
- Knowledge cutoff
- Jan 2025
- Released
- 2026-01
- Last updated
- 2026-01
← All NanoGPT listings