Anthropic Announced Sep 1, 2026

Claude Fable 5.1

Anthropic's most capable generally available model, sharing weights with the gated Claude Mythos 5.1, with unchanged list prices, cache reads cut 75%, and a nine-benchmark launch table against Fable 5, Opus 5, and GPT-5.6 Sol.

Context
1M
128K maximum output on the synchronous API
Effort
low → max
Five levels; the API defaults to high
Cache read
$0.25 / 1M
Down from $1.00 on Fable 5; batch API is 50% off
Knowledge cutoff
June 2026
Both reliable and training cutoffs per the model docs

Publisher pricing

Pricing below comes from the release announcement and does not represent a MiniRouter route or price.

Publisher-listed API price phases in US dollars per million tokens.
PhaseInput / 1MOutput / 1MBlended / 1MEffectiveSource
List$10.00$50.00$20.00Sep 1, 2026open-endedAnthropic

Publisher-reported benchmarks

Results are grouped by unit. Percentage, Elo, and task-count results never share a scale.

BenchmarkReported resultReported byAs ofSource
Terminal-Bench-Science 0.152.6%AnthropicSep 1, 2026Announcement ↗
Terminal-Bench 4.055.8%AnthropicSep 1, 2026Announcement ↗
CursorBench 3.2.073.4%AnthropicSep 1, 2026Announcement ↗
AutomationBench31.4%AnthropicSep 1, 2026Announcement ↗
OSWorld 2.0 · partial77.9%AnthropicSep 1, 2026Announcement ↗
OSWorld 2.0 · strict41.7%AnthropicSep 1, 2026Announcement ↗
Humanity's Last Exam · no tools60.9%AnthropicSep 1, 2026Announcement ↗
Humanity's Last Exam · with tools65.0%AnthropicSep 1, 2026Announcement ↗
BenchmarkReported resultReported byAs ofSource
GDPval-AA v21,853AnthropicSep 1, 2026Announcement ↗
  • Anthropic says Fable 5.1 and Mythos 5.1 share the same underlying model and pricing; Mythos 5.1 is limited to approved organizations with fewer safeguards.
  • Forced tool use returns an error, other models cannot read Fable 5.1 thinking blocks, and editing earlier turns invalidates thinking; see Anthropic's migration guide.
  • Anthropic ran the launch table with production safeguards on; tasks they intervened in scored zero. MiniRouter has not reproduced the figures.

Independent measurements

Artificial Analysis runs its own evaluations and timing against each model's first-party API. These are the only numbers on this page not reported by the publisher.

Claude Fable 5.1 (max with fallback) ranks #1 of 448 model families on their Intelligence Index. Their headline uses the highest-scoring effort setting; every listed setting is below.

Intelligence Index
65.7
Composite of their evaluations
Coding Index
81.6
Coding evaluations only
Blended price
$20
USD per 1M tokens, 3:1 input to output, first-party list
Output speed
69
Median tokens per second
First answer token
178s
Median seconds, including reasoning
SettingIntelligence IndexCoding IndexBlended priceOutput speedFirst answer token
Claude Fable 5.1 (max with fallback)65.781.6$2069178s
Claude Fable 5.1 (xhigh with fallback)64.880.7$206446s
Claude Fable 5.1 (high with fallback)62.579.1$20477.6s
Claude Fable 5.1 (medium with fallback)60.577.1$20475.1s
Claude Fable 5.1 (low with fallback)58.175.2$20505.7s

Source: Artificial Analysis, read Sep 2, 2026. Measured on the model's first-party API, not a MiniRouter route.

What is not independently confirmed

  • Every figure on this page is self-reported by the cited publisher. MiniRouter did not run these benchmarks.
  • Benchmark harnesses are not standardised. Results sharing a name across vendors have not been confirmed to use identical runs.
  • Publisher pricing can change, and introductory phases have stated end dates.
  • Claude Fable 5.1 is routable through MiniRouter today; live pricing on its model page is authoritative over the publisher figures here.

Related coverage

Other release dossiers and the current MiniRouter catalog.

Sources

  1. Anthropic announcement

    Published Sep 1, 2026

    Read at Anthropic
  2. Artificial Analysis model page

    Published Sep 2, 2026

    Read at Artificial Analysis

Editorial reference revised 2026-09-02. Catalog status comes from the generated routing snapshot.