Kimi
Kimi K2.7 Code
Quality Score49.2
Value Score38.4
Reliability44.9
Cache Discount80%
Model specification
- Reasoning
- reasoning
- Input modalities
- text, image, video
- Output modalities
- text
- Context window
- 256k tokens
- Weights
- open weights
Token prices USD per 1M tokens
- Input
- $0.950
- Output
- $4.00
- Cached Input
- $0.190
- Cached Output
- not available
- Cost per task
- $0.222
Capability
- Intelligence
- 43.0
- Coding
- 60.8
- Agentic
- 30.3
- Omniscience
- -10.2
- Correct
- 39.6%
- Blended price
- $0.723
Answer outcomes
Correct 39.6%Incorrect 49.8%Abstained 10.6%
Artificial Analysis benchmarks
- GDPval-AA v2
- 34%
- τ³-Banking
- 20%
- Terminal-Bench v2.1
- 67%
- SciCode
- 47%
- Humanity’s Last Exam
- 35%
- GPQA Diamond
- 90%
- CritPt
- 10%
- AA-Omniscience
- 45%
- AA-LCR
- 75%
Similar models
- Kimi K3 (low)KimiQuality Score 59.7Value Score 38.1
- Kimi K3 (max)KimiQuality Score 69.6Value Score 45.7
- Kimi Linear 48B A3B InstructKimiQuality Score not availableValue Score not available
- DeepSeek V4 Pro (Reasoning, High Effort)DeepSeekQuality Score 50.6Value Score 46.0
- MiMo-V2.5-ProXiaomiQuality Score 47.1Value Score 40.9
- GPT-5.6 Terra (low)OpenAIQuality Score 51.0Value Score 34.4