Stunning Terminal-Bench 4.0 chart maps every frontier agent model's score vs cost
iScienceLuvr · x · 2026-10-23
@iScienceLuvr shared Artificial Analysis' Terminal-Bench 4.0 scatter chart, commenting "the progress is staggering." The chart plots dozens of models — Claude Opus 5.5, GLM-5.3, GPT-6.1, Gemini, DeepSeek, Qwen, MiniMax and more — by agent benchmark score vs evaluation cost per task, with frontier models clustering in the 60-80% band and a dense field of budget and open-weight models below 40%.
More from Models
- Google Ultra users report Argon eating 20-30% of usage limit per prompt — Error_33_6070 · 2026-10-23
- Voice bot team: GPT Live 1 and Gemini 3.8 Thinking both fail in production — Mr_Nice_ · 2026-10-23
- KURE rerankers demoed: 149M nano model boosts the right answer to top spot — tomaarsen · 2026-10-23
- KURE-v3 debuts: first model suite covering all 4 Sentence Transformers types — tomaarsen · 2026-10-23
- Debate: are abliterated models causing more critical vulns, or just democratizing their discovery? — max_paperclips · 2026-10-23
- MiniMax 3.1 Flash runs locally on MLX as quant releases become the norm — pcuenq · 2026-10-23