Alex Getmanearly AI signals
EN
Alex Getman @alexgetmancom · · Updated

Tests & comparisons

Sonnet 5.5 Max even beats Opus 5.5 Max on Terminal-Bench, while staying close to Opus across the rest

Sonnet 5.5 Max even beats Opus 5.5 Max on Terminal-Bench, while staying close to Opus across the rest. And against GPT Sol, the price/performance gap looks pretty brutal.

Terminal-Bench 4.0 chart shows Sonnet 5.5 Max (blue line) surpassing Opus 5.5 Max (orange line) in accuracy at high cost, while GPT-5.6 Sol (gray line) performs significantly worse.
FrontierCode v1.1 chart shows Sonnet 5.5 (blue line) achieving high scores at low cost, but trailing Opus 5.5 (orange line) at higher levels, while GPT-6 Sol (gray line) shows moderate performance.
CursorBench 4.0 chart shows Opus 5.5 Max (orange line) leading at high cost, with Sonnet 5.5 Max (blue line) closely following, while GPT-5.6 Sol (gray line) lags in performance.
AA-Briefcase v1.1 chart shows Sonnet 5.5 Max (blue line) and Opus 5.5 Max (orange line) with similar results at high cost, while GPT-6 Sol (gray line) significantly underperforms in Elo.
AnthropicOpenAIClaudeGPT
Русская версияMarkdown

Analysis

Related posts