DataLearner logo

Muse Spark 1.1vsClaude Fable 5

Across 6 shared benchmarks, Claude Fable 5 leads overall: Muse Spark 1.1 wins 2, Claude Fable 5 wins 4, with 0 ties and an average score difference of -6.63.

Facebook AI研究实验室
Muse Spark 1.1

Facebook AI研究实验室 · 2026-07-09 · Reasoning model

Anthropic
Claude Fable 5

Anthropic · 2026-06-09 · Reasoning model

Muse Spark 1.12 wins(33%)(67%)4 winsClaude Fable 5

Benchmark scores

Grouped by capability, sorted by largest gap within each. 6 shared benchmarks.

AI Agent - Tool Usage

Claude Fable 5 2/3
BenchmarkMuse Spark 1.1Claude Fable 5Diff
Terminal-Bench 2.18019 / 44Thinking (With Tools)884 / 44Deep Thinking (With Tools)-8
MCP-Atlas88.101 / 38Thinking (With Tools)83.305 / 38Normal (With Tools)+4.80
OSWorld-Verified80.807 / 26Thinking (With Tools)851 / 26Thinking High (With Tools)-4.20

Coding and Software Engineer

Claude Fable 5 2/2
BenchmarkMuse Spark 1.1Claude Fable 5Diff
SWE-Bench Pro - Public61.5011 / 57Thinking (With Tools)80.301 / 57Deep Thinking (With Tools)-18.80
DeepSWE53.3017 / 27Thinking (With Tools)702 / 27Deep Thinking (With Tools)-16.70

General Knowledge

Muse Spark 1.1 1/1
BenchmarkMuse Spark 1.1Claude Fable 5Diff
HLE62.104 / 181Thinking (With Tools)595 / 181Deep Thinking (No Tools)+3.10

Specs

FieldMuse Spark 1.1Claude Fable 5
PublisherFacebook AI研究实验室Anthropic
Release date2026-07-092026-06-09
Model typeReasoning modelReasoning model
ArchitectureDenseDense
ParametersNot availableNot available
Context length1M1M
Max outputNot available128K

API pricing

Prices use DataLearner records when available; missing fields are not inferred.

ItemMuse Spark 1.1Claude Fable 5
Text inputNot public$10 / 1M tokens
Text outputNot public$50 / 1M tokens
Cache readNot public$1 / 1M tokens
Cache writeNot public$12.5 / 1M tokens

One or both models have incomplete public pricing.

Summary

  • Muse Spark 1.1leads in:General Knowledge (1/1)
  • Claude Fable 5leads in:AI Agent - Tool Usage (2/3), Coding and Software Engineer (2/2)

On average across the 6 shared benchmarks, Claude Fable 5 scores 6.63 higher.

Largest single-benchmark gap: SWE-Bench Pro - Public — Muse Spark 1.1 61.50 vs Claude Fable 5 80.30 (-18.80).

Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.