Kimi K2 Thinking
Thinking Kimi model for slower research passes, planning, and hard technical questions
- Context
- 262K
- Released
- 11mo ago
- Parameters
- —
- Input / 1M tokens
- $0.60
- Output / 1M tokens
- $2.50
- Cached input / 1M tokens
- $0.15
Sources:models.devOpenRouter
Benchmarks
DeepSWE is the primary coding benchmark where an exact model evaluation exists. Other sourced evaluations appear below; missing scores are not zero.
Not evaluated in DeepSWE 1.1 as of 2026-09-22. Results are only shown for exact model versions in the official dataset. Official leaderboard.
Other published benchmarks
| Evaluation | Reported score and source |
|---|---|
| FrontierMath-2025-02-28-Private | |
| FrontierMath-Tier-4-2025-07-01-Private | |
| Terminal Bench | 35.7% Terminal-Bench v2 Leaderboard · kimi-k2-thinking Agent: Terminus 2 Terminal-Bench v2 LeaderboardOriginal-source terms apply |
| Terminal Bench | 35.7% https://www.tbench.ai/leaderboard/terminal-bench/2.0 · kimi-k2-thinking Agent: Terminus 2 https://www.tbench.ai/leaderboard/terminal-bench/2.0Original-source terms apply |
Benchmark compilation: Epoch AI, Capabilities & benchmarking (CC BY attribution). Scores are converted to percentages and model names matched to the Vertical Tabs catalog. Individual evaluations are credited to their original sources. External results may carry separate licences and evaluation configurations; do not interpret scores from different harnesses as directly comparable.
Specifications
- Developer
- Moonshot AI
- Model family
- Kimi
- Type
- Language model
- Predecessor
- —
- Successor
- —
- Released
- November 6, 2025
- Model ID
moonshotai/kimi-k2-thinking- Openness
- Open weights
- License
- —
- Parameters
- —
- Input modalities
- text
- Output modalities
- text
- Modality
- text->text
- Tokenizer
- Other
- Max output tokens
- 98K
- Knowledge cutoff
- 2024-08
- Hugging Face
- moonshotai/Kimi-K2-Thinking
- Source freshness
- Metadata synced Oct 8, 2026, 10:17 PM UTC
Additional source evidence
- Source model
- moonshotai/kimi-k2-thinking
- Source family
- kimi-thinking
- Last updated
- 2025-11-06
- Knowledge cutoff
- 2024-08
- Model context
- 262K
- Model output limit
- 262K
- Open weights
- Yes
- File attachments
- No
- Reasoning
- Yes
- Tool calling
- Yes
- Temperature control
- Yes
Downloadable weights
Published benchmarks
| Benchmark | Score | Evaluation details | Evidence |
|---|---|---|---|
| SWE-Bench Verified | 71.3 · resolved | — | Source |
Complete source record (JSON)
{
"id": "moonshotai/kimi-k2-thinking",
"name": "Kimi K2 Thinking",
"description": "Thinking Kimi model for slower research passes, planning, and hard technical questions",
"family": "kimi-thinking",
"last_updated": "2025-11-06",
"knowledge": "2024-08",
"attachment": false,
"reasoning": true,
"tool_call": true,
"temperature": true,
"modalities": {
"input": [
"text"
],
"output": [
"text"
]
},
"limit": {
"context": 262144,
"output": 262144
},
"benchmarks": [
{
"name": "SWE-Bench Verified",
"score": 71.3,
"source": "https://huggingface.co/moonshotai/Kimi-K2-Thinking",
"metric": "resolved"
}
],
"release_date": "2025-11-06",
"open_weights": true,
"weights": [
{
"label": "Hugging Face",
"url": "https://huggingface.co/moonshotai/Kimi-K2-Thinking"
}
]
}Infrastructure
Provider availability
| Provider | Endpoint | Context | Input /M tokens | Output /M tokens | Cached input /M tokens | Cache write /M tokens | Uptime 1d | Data handling |
|---|---|---|---|---|---|---|---|---|
| google-vertex | 262K | $0.60 | $2.50 | — | — | 100.00% | ZDR: YesTraining: Unknown Prompt logging: Unknown Retention: Unknown Checked 2026-10-15 | |
| novita/bf16 | 262K | $0.60 | $2.50 | $0.15 | — | 100.00% | ZDR: YesTraining: Unknown Prompt logging: Unknown Retention: Unknown Checked 2026-10-15 |
Provider data:OpenRouter
Deployment graph
Can I run this?
Vertical Tabs only shows compatibility when it has structured evidence tying this exact model to hardware, runtime, quantization, memory use, and observed performance.