ModelsIndexCompareBest forBenchmarksStatusPricingAPI

Claude Fable 5

Anthropic·Operational·Released 2026-06-09textvision
Context1,000k
Max output128k
Input / 1M$10 ↗
Output / 1M$50 ↗
Cached in$1
Benchmarks44

Benchmarks

Cited public sources · rank is exact-benchmark, apples-to-apples

APEX-Agents (Mercor)#10/4963.6% ↗
Chess Puzzles (Epoch AI run)#16/9241% ↗
DeepSWE v1.1 (Datacurve)#5/2869.9% ↗
EBR-bench (Epoch AI run)#8/2439.5% ↗
FrontierMath Tier 4 v2 (Epoch AI run)#4/5790.2% ↗
FrontierMath Tiers 1-3 v2 (Epoch AI run)#9/7387% ↗
FrontierSWE V2 (Proximal Labs)#7/1847% ↗
Furniture Assembly (Epoch AI run)#17/2935.8% ↗
GPQA Diamond (Epoch AI run)#57/10285.9% ↗
LiveBench Agentic Coding (LiveBench)#7/5662.2% ↗
LiveBench Coding (LiveBench)#4/5686% ↗
LiveBench Data Analysis (LiveBench)#5/5680.5% ↗
LiveBench Instruction Following (LiveBench)#6/5675.8% ↗
LiveBench Language (LiveBench)#1/5690.7% ↗
LiveBench Mathematics (LiveBench)#8/5696% ↗
LiveBench Reasoning (LiveBench)#12/5689.7% ↗
MirrorCode (Epoch AI run)#3/963.9% ↗
Mystery Game Puzzles (Epoch AI run)#10/6852% ↗
OTIS Mock AIME 2024-2025 (Epoch AI run)#1/103100% ↗
SAGE (Vals AI)#9/7751.9% ↗
SimpleQA Verified (Epoch AI run)#6/7170.7% ↗
SWE-rebench 2026-05-15 to 2026-07-01 (Nebius)#1/1364.5% ↗
Terminal-Bench 4.0 (Vals AI)#9/4041.4% ↗
Vending-Bench 2 (Andon Labs)#22/595680.26 ↗
WeirdML (Håvard Tveit Ihle)#3/9291.9% ↗
LMArena Agent (LMArena)#8/480.0821 ↗
LMArena Elo#3/1411504.3 ↗
LMArena Vision (LMArena)#1/771308.8 ↗
LMArena WebDev (LMArena)#15/941625.3 ↗
GSO Opt@1 (GSO)#3/2776.5% ↗
tau2-bench Banking Knowledge (Sierra)#8/2239.7% ↗
OSWorld-Verified (XLANG)#1/1386% ↗
SimpleBench (SimpleBench)#6/7381.9% ↗
AutomationBench#4/417.4% ↗
Blueprint-Bench 238.6% ↗
CursorBench72.9% ↗
FrontierCode (Diamond)29.3% ↗
GDP.pdf29.8% ↗
Legal Agent Benchmark (Harvey's Held-Out Set)13.3% ↗
OfficeQA Pro#4/457.9% ↗
OSWorld-Verified#1/2085% ↗
SWE-Bench Pro#1/2180% ↗
SWE-bench Verified#1/5995% ↗
Terminal-Bench 2.1#10/2484.3% ↗

Compare

vs Gemini 4 Argonvs GPT-6.1 Solvs Claude Sonnet 5.5vs Muse Spark 1.3vs GPT-6 Solvs GPT-6 Lunavs Claude Opus 5.5vs GPT-6 Astra

More from Anthropic

Claude Sonnet 5.5Claude Opus 5.5Claude Fable 5.1Claude Opus 5Claude Sonnet 5Claude Haiku 4.5

Claude Fable 5 FAQ

What is Claude Fable 5?
Claude Fable 5 is a large language model from Anthropic, released on June 9, 2026. It accepts text and image input.
When was Claude Fable 5 released?
Anthropic released Claude Fable 5 on June 9, 2026.
How much does Claude Fable 5 cost?
Claude Fable 5 costs $10 per 1M input tokens and $50 per 1M output tokens through the Anthropic API. Cached input costs $1 per 1M tokens. Prices are from Anthropic's official pricing as of October 5, 2026.
What is Claude Fable 5's context window?
Claude Fable 5 has a 1M-token context window and can generate up to 128K output tokens.
How does Claude Fable 5 score on benchmarks?
Claude Fable 5 has 44 cited benchmark scores, including APEX-Agents (Mercor) 63.6%, Chess Puzzles (Epoch AI run) 41% and DeepSWE v1.1 (Datacurve) 69.9%. Each score links to its source.

Access Claude Fable 5 and every other model through one endpoint with automatic failover: Respan gateway.