M
Kimi K2 Thinking
Code source ouvertpar moonshotai · Sorti le 2025-11-06
38.1
score moyen
$0.47/1M
Prix d'entrée
$2.00/1M
Prix de sortie
131K tokens (~66 books)
Fenêtre de contexte
text
Type
Tested on 10 benchmarks with 38.1% average. Top scores: OTIS Mock AIME 2024-2025 (83.0%), GPQA diamond (79.0%), SWE-Bench Verified (Bash Only) (63.4%).
Scores de benchmark
| Benchmark | Catégorie | Score | Bar |
|---|---|---|---|
| OTIS Mock AIME 2024-2025 | math | 83.0 | |
| GPQA diamond | knowledge | 79.0 | |
| SWE-Bench Verified (Bash Only) | coding | 63.4 | |
| WeirdML | coding | 42.8 | |
| Terminal Bench | coding | 35.7 | |
| SimpleQA Verified | knowledge | 31.6 | |
| FrontierMath-2025-02-28-Private | math | 21.4 | |
| Chess Puzzles | knowledge | 20.0 | |
| APEX-Agents | agentic | 4.0 | |
| FrontierMath-Tier-4-2025-07-01-Private | math | 0.1 |