Claude Fable 5 on HealthBench Professional
rank 3 of 31 · updated September 30, 2026
Claude Fable 5 scores 0.660 on HealthBench Professional, rank 3 of 31 models on the board. Anthropic's top-tier model, positioned above the Opus line for the hardest reasoning work. HealthBench Professional scores models on 525 tasks drawn from real clinician conversations, graded criterion by criterion against physician-written rubrics on a 0 to 1 scale.
Result and API facts
| rank | 3 of 31 |
|---|---|
| score | 0.660 |
| lab | Anthropic |
| context window | 1.0M tokens |
| API price per 1M tokens | $10.00 in / $50.00 out |
| license | proprietary |
| source | Claude Fable 5 and Claude Mythos 5 System Card (system card) |
| released | 2026-06-09 |
Position in the field
The gap to the leader, GPT-6 Astra (Anthropic run) at 0.703, is 0.043. Directly above sits Claude Sonnet 5.5 at 0.692. Directly below sits Claude Opus 5.5 at 0.656. The scores on this page are compiled from published documents rather than from one controlled run, so a small gap between 2 models can reflect a difference in grader version, reasoning effort, or deployment setting as well as a difference in capability.
Source of this score
Read from Claude Fable 5 and Claude Mythos 5 System Card (system card, Anthropic, 2026-06-09). Vendor-reported. Confidence: verified. Configuration: length-adjusted, Anthropic protocol: adaptive thinking at max effort, Claude Opus 4.8 grader, averaged over 5 trials, no tools or custom system prompt (raw 70.3%). Measured as Claude Mythos 5; Anthropic itself prints 66.0 in the Fable 5 column of the Opus 5 card with footnote Mythos 5..
HealthBench Professional 66.0 - 64.7 56.9 51.8 -
p. 252, Table 8.1.A, row HealthBench Professional, column Mythos 5 (Fable 5 column '-'); Figure 8.18.2.A p. 298 bar label 66.0% on Claude Mythos 5 · 2 corroborating documents · full entry on the sources page
What does Claude Fable 5 score on HealthBench Professional?
Claude Fable 5 scores 0.660 on HealthBench Professional, which places it at rank 3 of 31 models on the board as of September 30, 2026. The number was read from Claude Fable 5 and Claude Mythos 5 System Card, listed on the sources page.
How much does Claude Fable 5 cost to run?
Claude Fable 5 is priced at $10.00 per million input tokens and $50.00 per million output tokens through Anthropic's API.
Head to head
Pairings with a dedicated comparison page are linked; every other difference is in the score-difference matrix.
- Claude Fable 5 vs GPT-6 Astra (Anthropic run)0.660 vs 0.703 · GPT-6 Astra (Anthropic run) by 0.043
- Claude Fable 5 vs Claude Sonnet 5.50.660 vs 0.692 · Claude Sonnet 5.5 by 0.032
- Claude Fable 5 vs Claude Opus 5.50.660 vs 0.656 · Claude Fable 5 by 0.004
- Claude Fable 5 vs GPT-6 Astra0.660 vs 0.647 · Claude Fable 5 by 0.013
- Claude Fable 5 vs Claude Fable 5 (September card)0.660 vs 0.633 · Claude Fable 5 by 0.027
- Claude Fable 5 vs Claude Fable 5.10.660 vs 0.621 · Claude Fable 5 by 0.039
- Claude Fable 5 vs GPT-6 Luna0.660 vs 0.608 · Claude Fable 5 by 0.052
- Claude Fable 5 vs GPT-6 Sol0.660 vs 0.608 · Claude Fable 5 by 0.052
- 0.660 vs 0.605 · Claude Fable 5 by 0.055
- 0.660 vs 0.598 · Claude Fable 5 by 0.062
- Claude Fable 5 vs Muse Spark 1.10.660 vs 0.593 · Claude Fable 5 by 0.067
- Claude Fable 5 vs Claude Sonnet 50.660 vs 0.578 · Claude Fable 5 by 0.082
- Claude Fable 5 vs GPT-5.6 Terra0.660 vs 0.577 · Claude Fable 5 by 0.083
- Claude Fable 5 vs Claude Opus 4.8 (Opus 4.8 grader)0.660 vs 0.574 · Claude Fable 5 by 0.086
- Claude Fable 5 vs Grok 4.70.660 vs 0.567 · Claude Fable 5 by 0.093
- Claude Fable 5 vs Claude Opus 4.80.660 vs 0.558 · Claude Fable 5 by 0.102
- Claude Fable 5 vs GPT-5.6 Luna0.660 vs 0.557 · Claude Fable 5 by 0.103
- Claude Fable 5 vs Muse Spark0.660 vs 0.541 · Claude Fable 5 by 0.119
- Claude Fable 5 vs GPT-5.6 Sol (August)0.660 vs 0.540 · Claude Fable 5 by 0.120
- Claude Fable 5 vs Claude Opus 4.70.660 vs 0.519 · Claude Fable 5 by 0.141
- Claude Fable 5 vs GPT-5.50.660 vs 0.518 · Claude Fable 5 by 0.142
- Claude Fable 5 vs Grok 4.60.660 vs 0.485 · Claude Fable 5 by 0.175
- Claude Fable 5 vs GPT-5.40.660 vs 0.481 · Claude Fable 5 by 0.179
- Claude Fable 5 vs GPT-50.660 vs 0.462 · Claude Fable 5 by 0.198
- Claude Fable 5 vs GPT-5.20.660 vs 0.459 · Claude Fable 5 by 0.201
- Claude Fable 5 vs Claude Sonnet 4.60.660 vs 0.442 · Claude Fable 5 by 0.218
- Claude Fable 5 vs GPT-5.6 Luna (August)0.660 vs 0.441 · Claude Fable 5 by 0.219
- Claude Fable 5 vs GPT-5.10.660 vs 0.396 · Claude Fable 5 by 0.264
- Claude Fable 5 vs GPT-5.5 Instant0.660 vs 0.384 · Claude Fable 5 by 0.276
- Claude Fable 5 vs MAI-Thinking-10.660 vs 0.350 · Claude Fable 5 by 0.310
Where the scores come from and how they are read is on the methodology page. The full ranking is on the leaderboard.