Skip to content

Benchmarks

Sources
8
Entries tracked
22,310
Last sync
27 min
All sources syncing normally
  • Arena · Text5 hours ago
  • Arena · WebDev5 hours ago
  • Arena · Vision5 hours ago
  • Arena · Agent5 hours ago
  • GPQA Diamond27 minutes ago
  • SWE-bench27 minutes ago
  • Terminal-Bench27 minutes ago
  • ARC-AGI-227 minutes ago

gpt-5.3-codex

74.8%±2.0%

72.8%CI76.8%

OpenAIOct 2026swing #28–#236

SWE-bench Verified ×17 of 32 models
ModelOrg± CISwingAdd to compare
2gpt-5.5-pre-releaseOpenAI——80.6%78.8%–82.4%#22–36
8gpt-5.4-2026-03-05OpenAISep 2026$2.50 / $1576.9%74.9%–78.8%#12–29
15gpt-5.3-codexOpenAIOct 2026$1.75 / $1474.8%72.8%–76.8%#28–236
17gpt-5.2-2025-12-11OpenAIDec 2025$1.75 / $1473.8%71.8%–75.8%#34–92
19gpt-5-2025-08-07OpenAIAug 2025$1.25 / $1073.6%71.5%–75.6%#64–149
25gpt-5.1-2025-11-13OpenAINov 2025$1.25 / $1068.0%65.9%–70.1%#45–242
26gpt-5-mini-2025-08-07OpenAIAug 2025$0.25 / $264.7%62.5%–66.8%#126–343

Showing SWE-bench Verified 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.