Skip to content

Benchmarks

Sources
8
Entries tracked
22,308
Last sync
9 min
All sources syncing normally
  • Arena · Text9 hours ago
  • Arena · WebDev9 hours ago
  • Arena · Vision9 hours ago
  • Arena · Agent9 hours ago
  • GPQA Diamond10 minutes ago
  • SWE-bench9 minutes ago
  • Terminal-Bench9 minutes ago
  • ARC-AGI-29 minutes ago

Kimi K2.7 Code

62±2

60interval64

moonshotJun 2026swing #58–#215

Kimi K2 0905

56±0

56interval57

kimiSep 2025swing #250–#250

Kimi K2

56±0

56interval57

kimiJul 2025swing #252–#252

Overall ×1WebDev Arena ×1Vision Arena ×1Factuality Arena ×1Search Arena ×1Document Arena ×1Agent Arena ×1GPQA Diamond ×1AIME (math) ×1FrontierMath ×1SimpleQA Verified ×1ARC-AGI-2 ×1Humanity's Last Exam ×1SimpleBench ×1LiveBench ×1BFCL v4 (function calling) ×1dormantSWE-bench Verified (live board) ×1Terminal-Bench 4.0 ×1MMLU-Pro (Artificial Analysis) ×1LiveCodeBench (Artificial Analysis) ×1IFBench (Artificial Analysis) ×1HHEM Hallucination (Vectara) ×14 of 1034 models
ModelOrgIntervalSwingiiiiiiiiiiiiiiiiiiiiiiAdd to compare
107Kimi K2.7 CodemoonshotJun 2026$0.67 / $3.3561.9 67.1Ranked 61.9 = (7 × 67.1 + 3 × 50) ÷ 10. 67.1 is the mean percentile over the 7 of 22 selected axes this model is measured on; fewer axes means a stronger pull toward 50.60–64427/22#58–215—58—————7682—39——6678—————72—
250Kimi K2 0905kimiSep 2025$0.60 / $2.5056.1 62.1Ranked 56.1 = (3 × 62.1 + 3 × 50) ÷ 6. 62.1 is the mean percentile over the 3 of 22 selected axes this model is measured on; fewer axes means a stronger pull toward 50.—333/22———————————————————786445—
252Kimi K2kimiJul 2025$0.57 / $2.3056.0 62.0Ranked 56.0 = (3 × 62.0 + 3 × 50) ÷ 6. 62.0 is the mean percentile over the 3 of 22 selected axes this model is measured on; fewer axes means a stronger pull toward 50.—363/22———————————————————816144—
446Kimi K2.5 (Thinking)moonshotJan 2026$0.50 / $2.5049.1 48.7Ranked 49.1 = (7 × 48.7 + 3 × 50) ÷ 10. 48.7 is the mean percentile over the 7 of 22 selected axes this model is measured on; fewer axes means a stronger pull toward 50.47–51667/22#70–446—50————————35466546——————8216

Blend is coverage-shrunk; the raw composite is shown greyed beside it · A low badge means fewer than 3 boards scored this model

The few things in AI and quantum that actually mattered this week — honestly summarised.

One email a week. Unsubscribe in one click. No tracking, no sharing.