Development

A throttled model scores ten points below itself and still leads

development2026-07-04

Claim

The re-released Fable 5 scored 54.8% on the APEX-SWE benchmark, ten points below its June self yet still nine clear of Opus 4.8, showing that even a throttled frontier outruns yesterday's best.

Reported metric: 54.8%, down 10, up 9 on rival

Reporting date: 2026-07-04 (not necessarily the event date)

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

Welcome to July 4, 2026 (theinnermostloop.substack.com)

Knowledge graphopen in browser →

Reported in

Welcome to July 4, 2026

Actor

Anthropic

Evidence for

Supersedes

Seventeen models have taken the lead, each reigning about seven weeks

Referenced by

Cite or correct this record

Citation formatCitation text

A throttled model scores ten points below itself and still leads
Innermost Loop KB — secondary record
Reporting date: 2026-07-04
KB permalink: https://nicholsn.github.io/innermost-loop-kb/developments/2026-07-04-a-throttled-frontier-still-outruns-yesterday
Claim: The re-released Fable 5 scored 54.8% on the APEX-SWE benchmark, ten points below its June self yet still nine clear of Opus 4.8, showing that even a throttled frontier outruns yesterday's best.
Attributed sources:
- Welcome to July 4, 2026 — https://theinnermostloop.substack.com/p/welcome-to-july-4-2026
This citation does not imply independent verification of linked-source claims.

Suggest a correction — opens a GitHub draft for you to review and submit.