Skip to content

Type: AI company or organization

SWE-Bench Pro appears in the radar stream as a company or organization. This page is a living index of dated mentions and sources — open Recent Updates and Backlinks for context, not a full company dossier.

Recent Updates

  • 2026-07-08: OpenAI says SWE-Bench Pro audit found task design issues that can distort results (OpenAI) · twitter.comLLM Evals OpenAI
  • 2026-07-08: OpenAI says SWE-Bench Pro was audited with investigator agents and five experienced engineers (OpenAI) · twitter.comAgents LLM Evals OpenAI
  • 2026-07-19: A postmortem on using multiple AI subscriptions to reduce research token burn (Hacker News) · quesma.comAgents Context Engineering Tool Use MCP Quesma Claude · Codex · Antigravity claude-mem Terminal-Bench · Artificial Analysis Claude Max 5x · Claude Fable 5 · Claude Opus 4.8 · Claude Sonnet 5 · GPT 5.5 · Gemini-3.1-Pro

FAQ

What is SWE-Bench Pro?

SWE-Bench Pro appears in the radar stream as a company or organization. This page is a living index of dated mentions and sources — open Recent Updates and Backlinks for context, not a full company dossier.

What does this page track?

Dated radar mentions, source links, related concepts, and builder-relevant context for SWE-Bench Pro, collected automatically by GROUNDING.

When was SWE-Bench Pro last mentioned?

SWE-Bench Pro was most recently mentioned in a radar update dated 2026-07-19.