Kimi-K2-Thinking - SWE-Bench Verified
SWE-Bench Verified resolved rate 63.4
View sourceMoonshot AI
Kimi K2 Thinking is the latest, most capable version of an open-source thinking model.
Kimi K2 Thinking is the latest, most capable version of an open-source thinking model.
56.0
Quality Score
---
Arena ELO
Undisclosed
Parameters
256K
Context
Sign in to join the discussion
4.8K
Downloads
0
Likes
May 2026
Released
Benchmarks
19
API
1
Recent launch, pricing, benchmark, and API signals linked to this model or its provider.
SWE-Bench Verified resolved rate 63.4
View sourceQuality: 40.9/100 | Price: $1.075/M tokens | Output: 119.446 tok/s | MMLU: 0.848% | HumanEval: 0.853%
View sourceQuality: 40.9/100 | Price: $1.075/M tokens | Output: 119.446 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 118.807 tok/s | MMLU: 0.848% | HumanEval: 0.853%
View sourceQuality: 40.9/100 | Price: $1.075/M tokens | Output: 114.826 tok/s | MMLU: 0.848% | HumanEval: 0.853%
View sourceQuality: 40.9/100 | Price: $1.075/M tokens | Output: 110.7 tok/s | MMLU: 0.848% | HumanEval: 0.853%
View sourceQuality: 40.9/100 | Price: $1.075/M tokens | Output: 118.807 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 114.826 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 110.7 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 107.768 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 106.603 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 111.061 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 83.735 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 84.162 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 62.672 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 42.84 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 44.205 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 43.038 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 90.299 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 84.441 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 98.887 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 99.304 tok/s | MMLU: 0.848% | HumanEval: 0.853%
Quality: 40.9/100 | Price: $1.075/M tokens | Output: 76.485 tok/s | MMLU: 0.848% | HumanEval: 0.853%
kimi-k2-thinking is now available through Ollama Cloud. 256K context window listed. Kimi K2 Thinking, Moonshot AI's best open-source thinking model.
SWE-Bench Verified resolved rate 63.4