Claude Fable 5

ModelActive

by Anthropic · family “claude-fable”

Claude Fable 5 is a Mythos-class model from Anthropic, built for autonomous knowledge work and coding. It supports text, image, and file inputs with text output, with reasoning support and... (description from the OpenRouter listing)

in: textin: imagein: fileout: textReasoningTool useStructured output

Epoch Capabilities Index

163.6 #3 of 268

90% CI 160.6 – 167.6

Listing ↗

Released

9 Jun 2026

Context window

1M tokens

Max output

128K tokens

Input $ / 1M tokens

$10.0

Output $ / 1M tokens

$50.0

Cached input $ / 1M

$1.00

Each benchmark shown separately with its own source

Benchmark results (21)

All benchmarks

Benchmark	Domain	Score	vs best recorded	Setting	Run	Source
Furniture Assembly	multimodal	35.8% ±6.2	43%	max	10 Sep 2026	Epoch ↗
LMCA	agents	61.2%	93%	high	—	External ↗
DTBench	reasoning	98.4%	100%	max	—	External ↗
Mystery Game Puzzles	games	52.0% ±5.0	62%	max	31 Jul 2026	Epoch ↗
EBR-bench	reasoning	39.5% ±2.0	52%	max	17 Jul 2026	Epoch ↗
MirrorCode	coding	63.9% ±10.4	87%	high	10 Aug 2026	Epoch ↗
Surface Evolver Bench	science	95.0%	100%	high	—	External ↗
FrontierMath-Tiers-1-3-v2-Private	math	87.0% ±2.0	93%	max	9 Jun 2026	Eval log ↗
FrontierMath-Tier-4-v2-Private	math	90.2% ±4.6	92%	max	9 Jun 2026	Eval log ↗
DeepSWE	coding	69.7%	94%	max	—	External ↗
PostTrainBench	agents	41.8%	100%	max	—	External ↗
ProofBench	math	95.0%	95%	max	—	External ↗
APEX-Agents	agents	63.6%	93%	—	—	External ↗
Chess Puzzles	games	41.0% ±4.9	57%	high	6 Aug 2026	Eval log ↗
Remote Labor Index	agents	16.1%	77%	—	—	External ↗
ARC-AGI-2	reasoning	89.2%	94%	max	—	External ↗
WeirdML	coding	91.9% ±0.0	98%	max	—	External ↗
OTIS Mock AIME 2024-2025	math	100.0%	100%	high	6 Aug 2026	Eval log ↗
SimpleBench	reasoning	81.9%	100%	max	—	External ↗
GPQA diamond	science	85.9% ±2.5	90%	max	6 Aug 2026	Eval log ↗
ARC-AGI	reasoning	98.5%	100%	max	—	External ↗

Source: Epoch AI Benchmarking Hub (CC BY 4.0). “External” rows are leaderboard results Epoch collects from third parties. Best reported setting per benchmark is shown.

API list price over time

Price history

$ per 1M tokens

4 observations (OpenRouter listing + Internet Archive snapshots). History accumulates with every ingest run; a single point means no change has been observed yet.