Skip to main content
Van Data Team

Current model · Anthropic

Claude Fable 5.2 benchmarks, pricing and specs

Claude Fable 5.2 is an Anthropic model released on November 9, 2026 that takes text and image input and produces text. It costs $8 per 1M input tokens and $40 per 1M output tokens and accepts up to 1,000,000 tokens of context. It scores 93.9% on GPQA Diamond, #12 of 43 current models (Epoch AI, measured November 10, 2026).

Per 1M input tokens
$8Listed today
Per 1M output tokens
$40Listed today
Context window
1M#28 of 70 current models
GPQA Diamond
93.9%#12 of 43 current models

Data verified November 10, 2026 · Dataset 2026-11-10

Share this

Specs

Key facts

Provider
Anthropic
Model family
Claude Fable
Release date
November 9, 2026
Status
Current
Access
API only
API model id
claude-fable-5-2
Context window
1,000,000 tokens
Max output
128,000 tokens
Input
text, image
Output
text
Task families
text, multimodal
Declared capabilities
reasoning

Charts

Against every current text model

Where Claude Fable 5.2 lands among the current models of its task, on every independent benchmark, on price and on context.

Where it stands

On GPQA Diamond, Claude Fable 5.2 ranks #12 of 43 current text models. 52 of 61 list a lower price per 1M input tokens. Each bar is the share of current models it beats.

Claude Fable 5.2 among the current text models, one row per benchmark, price and context window.

Independent benchmarks

GPQA Diamond

#12of 43

93.9% · ahead of 71%

FrontierMath Tiers 1-3 (v2)

#1of 43

95.1% · ahead of 100%

SimpleQA Verified

#1of 40

94% · ahead of 100%

Price and context

Price per 1M input tokens

#53of 61

$8 · cheaper than 12%

Price per 1M output tokens

#53of 61

$40 · cheaper than 12%

Context window

#28of 70

1M tokens · longer than 23%

Scores fromEpoch AI · 2026-11-10

Price

Claude Fable 5.2 pricing

It is the #8 most expensive of 61 current text models priced per 1M input tokens.

UnitPrice (USD)TierAppliesSource
1M input tokens$8standardCurrent
1M output tokens$40standardCurrent
1M input tokens$0.20cachedCurrent
1M input tokens$10cache write 5mCurrent
1M input tokens$16cache write 1hCurrent
1M input tokens$4batchCurrent
1M output tokens$20batchCurrent

Benchmarks

Independent benchmark results

Scores come from independent evaluators that ran the model themselves, pinned to a dated snapshot. Rank is among current models in the same snapshot. Anthropic's own launch figures are in the next section, because they are not measured the same way.

BenchmarkScoreRankMeasuredSource
GPQA Diamond93.9%#12 of 43November 10, 2026Epoch AI · 2026-11-10
FrontierMath Tiers 1-3 (v2)95.1%#1 of 43November 10, 2026Epoch AI · 2026-11-10
SimpleQA Verified94%#1 of 40November 10, 2026Epoch AI · 2026-11-10

Provider results

Reported by Anthropic

Anthropic's own results for Claude Fable 5.2, beside the rivals it chose, as its table lists them. Labs pick the benchmarks and run them their own way, so compare within one table only. Internal benchmarks and other evaluators' indexes are left out.

Claude Fable 5.2 launch table, November 9, 2026Anthropic · 2026-11-10

BenchmarkClaude Fable 5.2Claude Mythos 5.2Claude Fable 5.1Claude Opus 5.5GPT-6 AstraGemini 4 Carbon
FrontierCode 2.0best effort level per model49.1%Not listed32.4%45.5%42.8%41.1%
FrontierSWE 269.9%Not listed56.3%62.3%65.5%61.2%
AutomationBenchZapier51.4%Not listed31.4%42.5%41.4%53.9%
Terminal-Bench-Science 0.174.6%75.3%52.6%58.7%68.1%64.8%
WANDRrelaxed score82.1%Not listed68.7%72.3%Not listedNot listed
OSWorld 2.1offline set, partial credit89.3%Not listed83.5%87.2%79.5%81.5%
Remote Labor Index41.3%Not listed17.9%21.3%20.8%Not listed
Humanity's Last Examwith tools74.2%Not listed65.6%67.7%57.2%Not listed
Chartographyno tools73.4%Not listed44.8%64.4%71.0%72.5%

Anthropic's table from the Claude Fable 5.2 and Mythos 5.2 announcement, each model at its highest available effort, with Anthropic's safeguards on; competitor figures come from their developers' system cards or the benchmarks' leaderboards. On FrontierCode 2.0 Fable 5.2 and 5.1 ran at xhigh effort and Gemini 4 Carbon at medium. Anthropic gives a Mythos 5.2 score for Terminal-Bench-Science only. Left out here: Anthropic's own evaluations (CVE Rediscovery, de novo binder design, RoboDojo Sim), Irregular's CyScenarioBench, where each lab's score covers different tasks, and Artificial Analysis's AA-Briefcase.

Fit

Strengths, cautions and best uses

Strengths
  • Anthropic's most capable generally available model (November 9, 2026), for demanding reasoning and long-horizon agentic work, with a 1M-token context window and 128,000-token maximum output
  • Cheaper than Claude Fable 5.1: $8 input and $40 output per 1M tokens against $10 and $50, cache reads at $0.20 (0.025x input) and half price through the Batch API
  • Epoch AI measures 95.1% on FrontierMath Tiers 1-3 and 94% on SimpleQA Verified, the highest scores on both in this catalogue, and 93.9% on GPQA Diamond
  • In Anthropic's launch table it leads Claude Opus 5.5, GPT-6 Astra and Gemini 4 Carbon on FrontierCode 2.0 (49.1%), FrontierSWE v2 (69.9%) and OSWorld 2.1 (89.3%)
  • An optional fast mode generates up to 2.5 times faster at $16 input and $80 output per 1M tokens
Cautions
  • Gemini 4 Carbon scores higher on AutomationBench in Anthropic's own table (53.9% against 51.4%)
  • Safety classifiers block most penetration testing, exploit development and research biology requests and answer them with an older model (Claude Opus 4.8 for cybersecurity, Claude Opus 5 for biology); vetted organisations use Claude Mythos 5.2 for that work
  • Anthropic recommends starting most coding and knowledge work on Claude Opus 5.5, at half the price, and moving up to Fable only when it falls short; Arena has not rated Fable 5.2 yet
Best for
  • The hardest open-ended coding, research and agent runs, where Claude Opus 5.5 at high effort falls short
  • Advising or orchestrating cheaper Claude models in a multi-agent setup

From benchmark to production

Building on Claude Fable 5.2?

Public scores rarely hold on your own data. We test it against the alternatives on your real tasks, then build the agent or pipeline around the model that wins.

AI Model Cost AuditLaunch price $100, regular price $150What the audit includes

Book the cost auditAI agent development

Alternatives

Alternatives to Claude Fable 5.2

The closest current text models by price. The cheapest model that scores as well or better is listed first; failing that, the cheapest one within 3% of the benchmark's scale.

On score and price

Claude Fable 5.2 beside the 4 alternatives listed below, on GPQA Diamond and on blended price per 1M tokens. One of them costs less and scores as well or better.

Claude Fable 5.2 and its alternatives on score and price.

GPQA Diamond

Measured November 10, 2026, higher is better

Claude Fable 5.2
93.9%
Gemini 3.8 Flash
95.4%
GPT-6 Astra
95.8%
Claude Mythos 5.2
Not scored
Claude Opus 4.8
91%

Blended price per 1M tokens

Listed today, lower is better

Claude Fable 5.2
$16
Gemini 3.8 Flash
$1.50
GPT-6 Astra
$20
Claude Mythos 5.2
$16
Claude Opus 4.8
$10

Further reading

More on Claude Fable 5.2

Compare Claude Fable 5.2 with

All model comparisons

Articles about Claude Fable 5.2

Other Claude Fable models

FAQ

Claude Fable 5.2 FAQ

How much does Claude Fable 5.2 cost?

Claude Fable 5.2 costs $8 per 1M input tokens and $40 per 1M output tokens at the standard tier. It is the #8 most expensive of 61 current text models priced per 1M input tokens. Source: Anthropic, verified November 10, 2026.

What is the context window of Claude Fable 5.2?

Claude Fable 5.2 accepts up to 1,000,000 tokens of context. It can return up to 128,000 tokens in one response. That ranks #28 of 70 current text models by context length.

How does Claude Fable 5.2 score on independent benchmarks?

Claude Fable 5.2 scores 93.9% on GPQA Diamond, #12 of 43 current models (Epoch AI, measured November 10, 2026), 95.1% on FrontierMath Tiers 1-3 (v2), #1 of 43 current models (Epoch AI, measured November 10, 2026) and 94% on SimpleQA Verified, #1 of 40 current models (Epoch AI, measured November 10, 2026). These are independent evaluations, not Anthropic's own launch figures.

What benchmark results has Anthropic published for Claude Fable 5.2?

In the Claude Fable 5.2 launch table (November 9, 2026), Anthropic reports 49.1% on FrontierCode 2.0 (best effort level per model), 69.9% on FrontierSWE 2, 51.4% on AutomationBench (Zapier) and 74.6% on Terminal-Bench-Science 0.1. These are Anthropic's own figures, beside the rivals it chose, not independent measurements.

Is Claude Fable 5.2 open source or API only?

Claude Fable 5.2 is available through Anthropic's API only; its weights are not published.

What are the best alternatives to Claude Fable 5.2?

The closest current alternatives by price and measured performance are Gemini 3.8 Flash (91% cheaper per 1M input tokens, $0.75 vs $8), GPT-6 Astra (25% more per 1M input tokens, $10 vs $8), Claude Mythos 5.2 (same price per 1M input tokens, $8) and Claude Opus 4.8 (38% cheaper per 1M input tokens, $5 vs $8).

Sources

Sources

All 262 sources behind the benchmarks, with their versions, licenses and notes, are in the source catalog.

Cite this page

Van Data Team. "Claude Fable 5.2 Benchmarks, Pricing and Specs." AI Model Benchmark Dataset 2026-11-10. https://vandatateam.com/ai-model-benchmarks/claude-fable-5-2

Reuse the data: JSON · CSV · or explore every model in the AI model benchmark comparison.

Production evaluation

Deciding whether Claude Fable 5.2 fits your workload?

Public benchmarks narrow the shortlist. We test the finalists on your own tasks, data and cost limits, then build the routing and review controls to run the winner in production.

  • Evaluation cases and acceptance thresholds tied to real work
  • Model routing, fallbacks and cost controls
Compare all modelsBook a consultation