What are the verified coding and reasoning benchmarks for Anthropic: Claude Fable 5.1?

Anthropic: Claude Fable 5.1 achieves SWE-bench Verified: 76.2% (Extended Thinking), LiveCodeBench: 67.4%. Verified sources: Artificial Analysis Intelligence Index v4.3.2 (Artificial Analysis), Claude Fable 5.1 Architecture & Benchmark Card (Anthropic).

How does Anthropic: Claude Fable 5.1 compare to rival frontier coding models on SWE-bench and pricing?

Anthropic: Claude Fable 5.1 achieves SWE-bench Verified: 76.2% at $6/$30 per 1M tokens. Direct peer comparison: OpenAI GPT-6 Astra (82% SWE-bench, $6/$24); Muse Spark 1.3 (74.8% SWE-bench, $1.2/$6); xAI Grok 4.20 (58.5% SWE-bench, $2/$8); Claude Opus 5.5 (79.5% SWE-bench, $15/$75).

Does Swarm isolate Anthropic: Claude Fable 5.1 code edits from other agents?

Yes. Swarm provisions isolated Git worktrees for every session and subagent, preventing file race conditions and keeping your main working tree clean until verification passes.

Can Anthropic: Claude Fable 5.1 be assigned to specific subagent roles?

Yes. In ~/.swarm/config.json, you can configure Anthropic: Claude Fable 5.1 as Primary/Planner (DAG task decomposition), Coder (implementation in isolated worktrees), or Finder (read-only symbol search).

Anthropic: Claude Fable 5.1 — Specs, Context & Capabilities

Q: What is the recommended Swarm multi-agent topology for Anthropic: Claude Fable 5.1?

Swarm recommends Anthropic: Claude Fable 5.1 as Primary / Planner (Tier 1: Frontier Agentic Lead). Orchestrates multi-stage Task Programs with airtight boundary verification. Best paired with: deepseek-v4p1-flash or gemini-3-8-flash for high-speed subagent execution.

Q: How do I run Anthropic: Claude Fable 5.1 with Swarm?

Configure Anthropic: Claude Fable 5.1 via CLI using: swarm ctl model set --provider openrouter --model openrouter/anthropic/claude-fable-5.1, or switch models interactively inside the terminal TUI with /models.

Anthropic: Claude Fable 5.1 Benchmarks & Agentic Accuracy

Verified primary research & developer preference metrics

Primary Sources Verified

SWE-bench Verified

76.2%

Extended ThinkingAnthropic System Card ↗

LiveCodeBench

67.4%

ThinkingLiveCodeBench ↗

LMSYS Coding Elo

1435Elo

Coding ArenaLMSYS Arena ↗

Pricing ($ / 1M tokens)

$6/ $30

$0.6 cached readStandard API Pricing

What This Means in Real Software Development

Achieves an Intelligence Index of 53 on Artificial Analysis v4.3.2, second only to Claude Opus 5.5.
Scores 56 on the Finance & Accounting Index, excelling at complex quantitative workflows and multi-step spreadsheet deduction.
Delivers 65 TPS throughput at $7.63 weighted average cost per Intelligence task.
Engineered for deep multi-turn agentic sessions with zero instruction drift across large software systems.

Swarm Multi-Agent Topology

Tier 1: Frontier Agentic LeadRole: Primary / Planner

Best Paired With: deepseek-v4p1-flash or gemini-3-8-flash for high-speed subagent execution

Orchestrates multi-stage Task Programs with airtight boundary verification.

Git Worktree Isolation Advantage:

Directs parallel subagent worktrees while maintaining master branch integrity.

What Developers & Practitioners Are Saying

Real-world field observations, production strengths, and developer consensus

Field Consensus

"Celebrated as Anthropic's most versatile agentic flagship. Exceptional at complex tool calling, multi-file code editing, and holding context across 100+ turns."

Verified Real-World Strengths

Intelligence Index 53 with stellar multi-turn task completion
Resilient against edge-case hallucinations in complex codebases
Superior instruction obedience in long DAG pipelines

Production Caveats & Gaps

Higher cost per task ($7.63) than fast coder models
Output latency higher than subagent flash models

Primary Sources & Methodology Citations

Artificial Analysis Intelligence Index v4.3.2 — Artificial Analysis (September 2026)
Claude Fable 5.1 Architecture & Benchmark Card — Anthropic (May 2026)

Head-to-Head Frontier Comparisons: Anthropic: Claude Fable 5.1 vs Rivals

Standardized, side-by-side benchmark & economics comparison against peer models

Flagship Comparative Matrix

Model	Role / Tier	SWE-bench	LiveCodeBench	LMSYS Elo	AIME	Input / Output
Anthropic: Claude Fable 5.1OpenRouterCurrent	Tier 1: Frontier Agentic Lead	76.2%	67.4%	1435	—	$6 / $30
OpenAI GPT-6 AstraOpenAI Compare Anthropic: Claude Fable 5.1 vs OpenAI GPT-6 Astra →	Tier 1: Autonomous Chief Architect	82%	84.5%	1470	97%	$6 / $24
Muse Spark 1.3Meta Compare Anthropic: Claude Fable 5.1 vs Muse Spark 1.3 →	Tier 1: High-Speed Frontier Coder	74.8%	66.2%	1420	—	$1.2 / $6
xAI Grok 4.20xAI Compare Anthropic: Claude Fable 5.1 vs xAI Grok 4.20 →	Tier 1: Frontier Systems Specialist	58.5%	54.2%	1375	76%	$2 / $8
Claude Opus 5.5Anthropic Compare Anthropic: Claude Fable 5.1 vs Claude Opus 5.5 →	Tier 1: Master Principal Architect	79.5%	70.2%	1450	92%	$15 / $75

Explore All AI Model Comparisons Hub (Anthropic: Claude Fable 5.1 & Flagship Peers) →

Anthropic: Claude Fable 5.1 model overview

Anthropic: Claude Fable 5.1 is listed in the public Swarm catalog through OpenRouter. This page summarizes the verified catalog fields used by Swarm without claiming real-time access for every provider account.

Provider: OpenRouter
Family: OpenRouter
Context window: 1M tokens
Thinking support: low · medium · high · xhigh · max
Availability status: Available
Catalog ID: openrouter/anthropic/claude-fable-5.1

File input
Image input
Text input
Text output
Reasoning
Tools
Structured output
JSON

Run Anthropic: Claude Fable 5.1 in Swarm

Verified CLI workflow & multi-agent architecture

Quickstart Guide →

bash — swarm run

$swarm ctl model set --provider openrouter --model openrouter/anthropic/claude-fable-5.1

# Or switch interactively inside the terminal TUI:

>/models

Subagent Role Matrix

Assign Anthropic: Claude Fable 5.1 to specialized system roles in ~/.swarm/config.json:

Primary / Planner: Orchestrates full task DAGs and multi-stage pipelines.
Coder Subagent: Generates code and unit tests inside private Git worktrees.
Finder Subagent: Performs read-only code search, symbol discovery, and mapping.

System Agent Roles →Why Swarm Architecture →