Claude Fable 5.2
by Anthropic
Today's price · per 1M tokens
Input
$8.00
per 1M tokens
Output
$40.00
per 1M tokens
Blended
$16.00
blended $/1M — 3:1 weighted input:output
Cached input
$0.200
3% of input — prompt caching
How this price is scoped: Cache hits and refreshes on Claude Fable 5.2 (and Claude Mythos 5.2) are priced at 0.025x the base input price — $0.20 per 1M tokens. Base $8/$40 per 1M tokens; Batch API $4/$20; fast mode $16/$80. Verified 2026-11-09 against platform.claude.com/docs/en/about-claude/pricing.
Price receipt
We read Anthropic's own pricing page on and found Claude Fable 5.2 listed with a price on the same row — that read is where the number above comes from. We keep the page text we read; its content hash is 5955d90190.
Overview
Anthropic's newest Fable flagship for demanding reasoning and long-horizon agentic work — vision and tool use, with a 1M-token context window and up to 128K output tokens. $8/$40 per 1M tokens, 20% below Claude Fable 5.1, with cache hits at $0.20 per 1M tokens (0.025x input).
Capabilities
struck through = not supportedBenchmark performance
accuracy % · higher is betterWe hold no per-benchmark accuracy scores for this model yet, so it has no accuracy average. That is a gap in our coverage, not a sign the model is untested — independent evaluators often publish a composite index for a new model long before releasing its per-benchmark numbers. The percentile above is its standing across the independent composites below.
- AA Intelligence Index: 64.6 (max effort) (Artificial Analysis composite across reasoning, knowledge and coding evals)
- AA-Omniscience Index: 66.8 (max effort) (−100–100)
- GDPval-AA v2: 1831.1 ELO (Real-world work tasks, human baseline = 1000)
- Arena WebDev: 1837.8 ELO (Human preference on building web apps)
- Arena Agent: 0.304 (Score from real agent sessions (IPS); higher is better)
- Arena Agent cost per task: $7.43 (Median measured cost of one real Agent Arena task)
- Epoch Capabilities Index: 173.4 (CI 169.9–177.5) (Epoch AI's index fitted across many benchmarks)
- FrontierMath Tiers 1–3: 95.1% (max effort) (Research-level maths (funded by OpenAI, run by Epoch); best run across the effort settings Epoch tested)
- FrontierMath Tier 4: 100% (max effort) (Hardest research maths; best run across the effort settings Epoch tested)
Source: Artificial Analysis (artificialanalysis.ai)
- Humanity's Last Exam: 74.3% (max effort)
- AA-LCR (long-context reasoning): 83.3% (max effort)
- CritPt (research-level physics): 35.4% (max effort)
- Terminal-Bench 4.0: 59.6% (max effort)
- SciCode: 67.7% (max effort)
- Cost to run the AA Intelligence Index: $26.54 per task, 45.6 min per task (max effort)
- Measured by AA: 100.1 tokens/sec output, 258.83s to first token (max effort)
- Ranks #1 of 189 comparably-measured models by percentile score, across 3 independent measurements
- Ranks #178 of 189 comparably-tested models by normalized performance per dollar
100.1 tokens/sec output 258.83s latency to first token (TTFT)
Throughput: Artificial Analysis, measured 2026-11-14 · Source: Artificial Analysis (artificialanalysis.ai)
Source: Artificial Analysis · updated · See full rankings →
Specifications
- Provider
- Anthropic
- Context window
- 1M tokens
- Modality
- text, image
- Parameters
- Proprietary
- Open source
- No — proprietary
- Released
- Status
- Current
- Last updated
- Tags
Availability verified: — listed on Anthropic's own page