Ruff author Charlie Marsh: GPT-6.2 is a huge step-up, cost-score curves are 'insane'
charliermarsh · x · 2026-11-18
Charlie Marsh (ruff/uv) calls GPT-6.2 an enormous step-up for his own work and shares OpenAI's Agents' Last Exam v1.1 chart: GPT-6.2 Astra hits 78% at $2-5 per task, well ahead of Claude Fable 5.2 (73% at $60/task) and Claude Opus 5.5 (71% at $15/task), with Gemini 4 Carbon close behind at low cost.
Related event: Ruff creator Charlie Marsh calls GPT-6.2 a huge leap for his work(3 posts)→
More from Models
- Head-to-head test of three flagship models: Fable 5.2 still leads, but not a clean sweep — petergyang · 2026-11-18
- Asking Gemini 10,701 questions: Google Search rewrites ~45% of brand picks — dejanseo · 2026-11-18
- Cursor's CursorBench 4.0: Gemini 4 Carbon tops at 60.4%, DeepSeek under $1 per task — mattyp · 2026-11-18
- Salesforce's SOV-Transcribe-1 ASR model trends on Hugging Face — salesforce-sovereign-intelligence · 2026-11-18
- Test: Google's AI Mode default Gemini 4 Hydrogen knows almost nothing about brands without search — dejanseo · 2026-11-18
- OpenAI explains the 6.2 influencer/community vibe gap: demo videos used pre-release models — i_dg23 · 2026-11-18