Modelpeers
GPT-6 Astra characterClaude Sonnet 5.5 characterClaude Opus 5.5 character

Claude Opus 5.5, Sonnet 5.5, GPT-6 Astra

6 models from Anthropic and OpenAI and Google

Benchmark

Vals AI Terminal-Bench 4.0 with one harness: Opus 5.5 65.2%, Astra 59.6%, Argon 57.6%

Running every model in the same Mini-SWE-agent harness, Vals AI reports Claude Opus 5.5 at 65.15%, Sonnet 5.5 64.14%, GPT-6 Astra 59.60%, Gemini 4 Argon 57.58%, GPT-6.1 Sol 55.05% and Opus 5 53.53%. GPT-6.1 Sol cost $1.72 per test against $13.20 for Opus 5.5.

Source

"As of October 19, 2026, Claude Opus 5.5 (Mini-SWE-agent) ranks first on Terminal-Bench 4.0 with 65.15%"

Vals AI, Terminal-Bench 4.0 leaderboard

Published

Open the sourcevals.ai

Modelpeers tracks changes to AI agents and models, each from its source. Follow Claude Opus 5.5 to get a weekly email when it changes.