Skip to content

Benchmarks

Sources
8
Entries tracked
22,308
Last sync
25 min
All sources syncing normally
  • Arena · Text10 hours ago
  • Arena · WebDev10 hours ago
  • Arena · Vision10 hours ago
  • Arena · Agent10 hours ago
  • GPQA Diamond25 minutes ago
  • SWE-bench25 minutes ago
  • Terminal-Bench25 minutes ago
  • ARC-AGI-225 minutes ago

deepseek-v3.2

1,425±3

1,421CI1,428

deepseekDec 2025swing #86–#247

Overall ×110 of 396 models
ModelOrg± CISwingAdd to compare
110deepseek-v3.2-exp-thinkingdeepseekOct 2026—1,4251,419–1,432#77–180
111deepseek-v3.2deepseekDec 2025$0.28 / $0.421,4251,421–1,428#86–247
115deepseek-v3.2-thinkingdeepseekOct 2026—1,4221,419–1,426#103–345
117deepseek-v3.2-expdeepseekSep 2025$0.27 / $0.411,4221,415–1,428#92–334
123deepseek-v3.1deepseekAug 2025—1,4171,411–1,423#101–158
125deepseek-v3.1-terminus-thinkingdeepseekOct 2026—1,4161,406–1,426#97–137
127deepseek-v3.1-thinkingdeepseekOct 2026—1,4161,409–1,422#106–291
129deepseek-v3.1-terminusdeepseekSep 2025$0.27 / $11,4151,405–1,424#55–168
154deepseek-v3-0324deepseekMar 2025—1,3951,391–1,399#137–401
192deepseek-v3deepseekDec 2024—1,3581,354–1,363#178–644

Showing Overall 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.