Skip to content

Benchmarks

Sources
8
Entries tracked
22,318
Last sync
50 min
All sources syncing normally
  • Arena · Text1 hour ago
  • Arena · WebDev1 hour ago
  • Arena · Vision1 hour ago
  • Arena · Agent1 hour ago
  • GPQA Diamond50 minutes ago
  • SWE-bench50 minutes ago
  • Terminal-Bench50 minutes ago
  • ARC-AGI-250 minutes ago

DeepSeek-R1

53.3%±7.5%

45.8%CI60.9%

DeepSeekJan 2025swing #117–#389

AIME (math) ×17 of 200 models
ModelOrg± CISwingAdd to compare
107DeepSeek-R1-0528DeepSeekMay 2025$0.50 / $2.1566.4%61.5%–71.3%#89–250
120DeepSeek-R1-Distill-Qwen-32BDeepSeek——55.6%49.4%–61.8%#255–778
122DeepSeek-R1DeepSeekJan 2025$0.70 / $2.5053.3%45.8%–60.9%#117–389
124DeepSeek-R1-Distill-Llama-70BDeepSeekJan 2025$0.80 / $0.8051.4%45.3%–57.4%#286–643
125DeepSeek-R1-Distill-Qwen-14BDeepSeek——50.6%44.6%–56.5%#347–812
129deepseek-r1-0528-qwen3-8b———43.9%37.1%–50.7%#465–920
142DeepSeek-R1-Distill-Qwen-1.5BDeepSeek——21.4%16.1%–26.7%#407–1029

Showing AIME (math) exactly as the source published it — its own score and confidence interval, no blending. Select a second axis to build a weighted composite.

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.