Claude Fable 5.2 on BenchGecko · Anthropic's new top model at $8 / $40
Anthropic released Claude Fable 5.2 and Claude Mythos 5.2 on 2026-11-09. BenchGecko lists Fable 5.2 at $8 input and $40 output per M tokens with a 1M token context.
What launched
Anthropic released Claude Fable 5.2 on November 9, 2026, alongside Claude Mythos 5.2, its cybersecurity focused sibling. Fable 5.2 is generally available on the Claude platform and through major clouds. Mythos 5.2 is limited to vetted security professionals through Anthropic's Cyber Verification Program and to Enterprise customers through Claude Security.
Specs
From Anthropic's model documentation:
- Context · 1M tokens (default and maximum)
- Max output · 128K tokens
- Inputs · text and images · output text
- Thinking · adaptive, always on, steered by the effort parameter (default high)
- Knowledge cutoff · June 2026
Pricing on BenchGecko
Both models list at $8 input and $40 output per million tokens, with Fast Mode at $16 and $80. Prompt cache reads cost 2.5% of the base input price. For comparison on the same docs page, Claude Opus 5.5 lists at $4 / $20 and Claude Sonnet 5.5 at $2 / $10, so Fable 5.2 is twice the price of Opus 5.5.
Vendor reported benchmarks
Anthropic reports these figures, at each model's highest available effort with safeguards on. BenchGecko records vendor claims as published.
- FrontierCode 2.0 · 49.1% (Fable 5.1 32.4% · Opus 5.5 45.5% · GPT-6 Astra 42.8% · Gemini 4 Carbon 41.1%)
- FrontierSWE v2 · 69.9% (Fable 5.1 56.3% · GPT-6 Astra 65.5%)
- Terminal-Bench-Science 0.1 · 74.6% (Mythos 5.2 75.3%)
- OSWorld 2.1 (offline set, partial) · 89.3%
- Humanity's Last Exam (with tools) · 74.2%
- Remote Labor Index · 41.3%
- AutomationBench · 51.4% (Gemini 4 Carbon 53.9%)
- CVE Rediscovery · 89.6% (Mythos 5.2)
Sources
Source: BenchGecko
Written automatically from BenchGecko's tracked data for this period; every number comes from the changelog, model, mindshare and pricing datasets. Checked against the source data before publishing.