Source record

Published by Amp News (ampcode.com) (T1 source). The original is at https://ampcode.com/news/opus-5.5.

Pipeline notes

The summary and note below are generated by the signal pipeline — they are Beyond Desk’s reading, not quotations from the source.

Amp replaced GPT-5.6 Sol with Claude Opus 5.5 as the default model in its coding agent's medium mode after internal evals showed Opus 5.5 solved 65% of tasks versus 61% for GPT-5.6 Sol, at 10% lower cost. Opus 5.5 is more steerable, accepting mid-work corrections, and is capped at high reasoning effort. Medium is recommended for most work; ultra with Fable 5.1 is reserved for high-uncertainty tasks such as an enterprise SSO email migration.

Firsthand benchmark numbers, a steerability pattern, and a concrete medium-vs-ultra decision rule, all from the vendor's own evals with cited task scores and costs.

Cited by

No citations on record.