The evidence wall

Every claim, with its receipt

All launch-day numbers in one place, each traceable to a source. Third-party evals get added as they publish — claims without receipts don't get on the wall.

Guide current as of June 11, 2026. Fable 5 is included on Pro/Max/Team plans through June 22; usage credits apply after. Latest news →

ClaudeFieldGuide › Benchmarks

Benchmark table comparing Claude Fable and Mythos to other leading models
The master table from Anthropic's announcement — Fable and Mythos against other frontier models.

Software engineering

ClaimSource
50-million-line Ruby migration completed in one day (team scoped two months)Stripe, via announcement
Highest frontier score on FrontierCode — at medium effort (chart, chart 2)Cognition
"State of the art model on CursorBench"Michael Truell, Cursor
Long-horizon autonomy "exceeded previous benchmarks"Mario Rodriguez, GitHub
Highest on ViBench end-to-end vibe-coding, "nearly saturating base use cases"Michele Catasta, Vibe

Knowledge work & research

ClaimSource
First past 90% on core analytics benchmark — 10 points over Opus; top of Finance BenchmarkIzzy Miller, Hebbia
"Aced trading-analysis evaluations nearly across the board"IMC
Strongest on frontier physics at a third of the reasoning tokens; in 36 hours reached "nearly" where GPT-5.5 landed in four days (partner report, not an independent eval)Matthew Pines, Notation Capital — via announcement, June 9, 2026

Vision, memory, autonomy

ClaimSource
Completed Pokémon FireRed vision-only with a minimal harnessAnnouncement
Rebuilds web-app source from screenshots; extracts precise values from scientific figuresAnnouncement
With file memory in Slay the Spire: improved 3× more than Opus 4.8; reached final act 3× more oftenAnnouncement

Independent results (early — day two)

ClaimSource
#2 of 123 models, 96/100 overall, on the provisional leaderboard; top-tier coding and agentic scores; #18 on multimodal/grounded (79) — the first independent wrinkle in the vision storyBenchLM
95.0% on SWE-Bench VerifiedReported via BenchLM aggregation
AA-Omniscience knowledge benchmark: 40 — seven points above the previous leader, Gemini 3.1 ProArtificial Analysis, via Simon Willison
91/100 on Every's internal senior-engineer exam (Opus 4.8: 63)Every, via AI+ Founders
Guardrail fallback observed on ~8–9% of tasks (mostly scientific) — vs the announced <5% session average; methodology differs (tasks vs sessions), flagged as the launch's most checkable open questionThe Decoder
Sustained 12-hour autonomous runsDigital Today

Early and provisional — leaderboards move in week one. Tracked in News.

Mythos research results

Run with safeguards lifted, by vetted partners — context on the Mythos page:

ClaimSource
~10× acceleration on aspects of drug design; 9 of 14 protein targets yielded strong candidates (chart)Announcement
Hypotheses preferred ~80% over Opus-class in blinded comparisons; one corroborated independentlyAnnouncement
Genomics model beat a recent Science publication at 1/100th the size, from a week-long autonomous runAnnouncement
Beat dedicated protein language models on AAV prediction (chart)Announcement
Reading the wall: everything above is launch-day material — Anthropic's own numbers and partner quotes from the announcement. Independent third-party evals get their own section here as they publish. If a claim has no receipt, it doesn't go up.
Live