Bug Hunt leaderboard: GPT-6.2 Astra Max tops at 62.7/105 while Sol Max fixes 54.7 bugs for just $7
PawelHuryn · x · 2026-11-19
Pawel Huryn shares the Bug Hunt Bench leaderboard breaking down GPT-6.2 performance across effort levels (105 planted bugs, blind-verified fixes only).
- GPT-6.2 Astra Max leads at 62.7/105 (xhigh: 60), but takes 121.8 min and $28.35 per run
- GPT-6.2 Sol Max scores 54.7 for only $7.00 in 68 min — the standout value pick; Sol spans 36 ($1.96, low) to 54.7 (max)
- Anthropic's Fable 5.2 Max hits 58 but costs $139.65, nearly 5x Astra Max; Opus 5.5 trails on both axes (low: 22.3)
Huryn argues OpenAI keeps improving token efficiency each generation, making it the best value even after subscription multipliers. His workflow: delegate all implementation and review to Sol (6.1), use Opus 5.5 for planning, coordination, and judgment. GPT-6.2 Astra high/medium/low tiers are still rolling out.
Related event: GPT-6.2 Astra Tops Bug Hunt Bench; Sol Matches Rivals at Fraction of Cost(5 posts)→
More from coding & agent
- Google Labs launches Stitch Loop, an always-on product agent for post-launch improvement — GoogleLabs · 2026-11-19
- Running untrusted agents on GKE: sandboxing, cold starts, and state — rseroter · 2026-11-19
- Jane Street opens first Formal Methods internship to validate agent-generated code — hargup13 · 2026-11-19
- Malicious GitHub repos trick Claude, Codex and Copilot into silently running a credential stealer — ccerrato147 · 2026-11-19
- Gemini 4 Carbon's slowness is a trajectory problem—stop conditions cut runtime 37→28 min — petrusenko_max · 2026-11-19
- Running a Faceless Instagram Page with a Team of Sparks: One Developer's Pipeline Plan — abhi9u · 2026-11-19