HyperAI
Command Palette
Search for a command to run...
kimi-k2-thinking
Detailed information about kimi-k2-thinking AI model
At a glance
Modalities
Context window
262,144
Pricing
$0.60$2.50
Capabilities
Tool Calling
Connect the model to external tools and systems.
Reasoning
The model thinks before responding.
Benchmarks
| Benchmark | Score | Link |
|---|---|---|
swe-bench / bash-only | 63.4% | moonshotai.github.io |
swe-bench / Verified | 63.4% | mini-swe-agent.com |
Pricing Details
Input
Tokens
$0.60/1M tokens
Cache read
$0.15/1M tokens
Output
Tokens
$2.50/1M tokens
Model Owner
Available Providers
OpenRouter
openrouter/kimi-k2-thinking
Released on 11/6/2025
Want to try more models?
Deploy on HyperAI with one click and experience model performance instantly.