GPT-6.1 Sol already takes half the cells in the speed-vs-cost intelligence chart
randal_olson · x · 2026-10-01
- Follow-up on Olson's Artificial Analysis chart: each cell is the best Intelligence Index score within both limits; reply time is AA's end-to-end time for a 500-token response, cost is per Index task.
- GPT-6.1 Sol, released this week, already takes about half the cells.
- The underlying tool, evident-charts, is open source: it teaches AI coding agents (Claude Code, Codex, Cursor, etc.) to make clear, honest charts and checks each one — first vetting the data (mixed totals, duplicate rows, placeholder codes, preliminary months), then critiquing and rebuilding bad chart drafts.
Related event: Data Visualization Reveals Best-Value Smart AI Models(2 posts)→
More from coding & agent
- Overmind open-sources a platform that fine-tunes agents from production traces — kimmonismus · 2026-10-01
- Dev turns Cloudflare birthday-week blog flood into Claude prompts — threepointone · 2026-10-01
- Fable 5.1 vibe check: stronger coding, half the tokens of Opus 5, and it finally talks like a person — every · 2026-10-01
- Zapier CEO grades three levels of AI-powered PM work live, top level is a full agent operating model — aakashgupta · 2026-10-01
- Dev Benchmark: Claude Outdelivers Codex on Same Task, Overnight PR vs Getting Stuck — QuixiAI · 2026-10-01
- Agentic AI Is Nothing New: Multi-Agent Systems Plus an Orchestrator — DavidLinthicum · 2026-10-01