eric provencher(@pvncher):This is cool, but the cost per task averages under $0.25 What tasks have you given a model lately that costs less that $1? Piecemeal evals are saturated. Agents run for hours now.

eric provencher

@pvncher

Codex DX @Openai | built @repoprompt | prev XR @unity

Joined June 2013

4.9K Following 45.3K Followers

eric provencher@pvncher

2026.09.23 12:01

This is cool, but the cost per task averages under $0.25 What tasks have you given a model lately that costs less that $1? Piecemeal evals are saturated. Agents run for hours now.

Next.js@nextjs

2026.09.22 21:52

GPT-6 Sol debuts at 97% on Next.js evals, tying the highest success rate. Claude Opus 5.5 also reached 97%, joining Claude Fable 5.1. Opus ranks first among the three on average cost. View the leaderboard ↓