You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Add Kimi K2 thinking models to Moonshot AI and OpenRouter
- Add kimi-k2-thinking and kimi-k2-thinking-turbo to moonshotai provider
- Add moonshotai/kimi-k2-thinking to openrouter provider
- All models marked as reasoning=true for thinking capabilities
- Set open_weights=true to match other Kimi K2 models
- Context window: 256K (262,144 tokens)
- Max output: 256K (262,144 tokens)
- Knowledge cutoff: 2024-08
- Cache read: $0.15/M tokens for all models
- Pricing: k2-thinking ($0.60 input, $2.50 output), k2-thinking-turbo ($1.15 input, $8 output)
- Released November 6, 2025
0 commit comments