Compare / Claude Fable 5.1 vs Muse Spark 1.3

Claude Fable 5.1 vs Muse Spark 1.3

Which is cheaper, which is more capable, and which does better at each kind of work, from 82 results on 7 sources that measured both models.

AnthropicClaude Fable 5.1 Released 1 Sep 2026 · $10.00 in · $50.00 out
MetaMuse Spark 1.3 Released 2 Sep 2026 · $1.25 in · $4.25 out
Claude Fable 5.1 wins 3723 too close to callMuse Spark 1.3 wins 22

Of the 82 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.

Results as of 21 November 2026 (catalogue release 2026-11-21-af3a2faf6a5c). See also Anthropic vs Meta, every model from both labs.

Claude Fable 5.1 or Muse Spark 1.3? The short answer

Pick Claude Fable 5.1 if quality matters most: it scores higher (Spring Prompt overall 64 against 59). Pick Muse Spark 1.3 if cost matters more: it costs 6.0 times less per product listing on CatalogBench. Muse Spark 1.3 is also faster.

  • Spring Prompt overall: Claude Fable 5.1 64, Muse Spark 1.3 59 (out of 100; how the score works).
  • Muse Spark 1.3 is cheaper to run: $0.016 against $0.099 per product listing on CatalogBench (6.0 times less).
  • Claude Fable 5.1 scores higher on the Artificial Analysis Intelligence Index: 53.4 against 48.1.
  • Muse Spark 1.3 writes faster: 254 tokens a second against 72.
  • Muse Spark 1.3 is newer: released 2 Sep 2026, against 1 Sep 2026 for Claude Fable 5.1.
  • On our own benchmarks, on the results both models have: Claude Fable 5.1 does better for decks from an analysis and marketing planning; Muse Spark 1.3 for product listings.
  • On third-party benchmarks, on the results both models have: Claude Fable 5.1 does better for agents and tool use, professional work, reasoning and knowledge and sticking to the facts.

Head to head

64
Spring Prompt overall
59
53.4
Intelligence Index
48.1
$20.00
Price per million tokens, blendedlower is better
$2.00
72
Output speed, tokens a second
254
1,000,000
Context window, tokens
1,048,576

Every fact side by side

Claude Fable 5.1Muse Spark 1.3
Price per million tokens$10.00 in · $50.00 out$1.25 in · $4.25 out
Blended (3 in : 1 out)$20.00$2.00
Cheapest host, blended$20.00 (Amazon Bedrock)–
Cost per product listing (CatalogBench)$0.099$0.016
Cached input, per million$0.25 cached input$0.15 cached input
Intelligence Index53.448.1
Spring Prompt overall6459
Output speed, tokens a second72254
Context window1,000,000 tokens1,048,576 tokens
Released1 Sep 20262 Sep 2026 (newer)
WeightsClosedClosed
Developer based inthe USthe US

Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.

Which is better for what

Each model's rank among the models on sale today, for each kind of work, from our own benchmarks and licensed sources. Further right is better; the better rank is in green.

Product listingsTurning a sparse product feed and photos into listings that can go live 27 of 36 20 of 36
Decks from an analysisTurning a finished analysis into a deck you could present as it is 11 of 36 28 of 36
Marketing planningPlanning a year of ad spend without overspending 12 of 36 28 of 36
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls 19 of 159 13 of 159
Professional workReal tasks from banking, consulting and law, business documents and freelance projects 4 of 51 8 of 51
Reasoning and knowledgeHard questions across science, maths and general knowledge 5 of 187 11 of 187
WritingWhat people prefer in blind comparisons, and judged writing quality 19 of 169 7 of 169
Sticking to the factsSummarising without inventing things, and factual answers 5 of 139 11 of 139
Speed under pressureGood decisions against a real clock (fast chess) 39 of 39 36 of 38

Their slides, side by side

The same brief for both: Manchester store investment review for Hearthside Coffee, an invented company. First slide of each deck from DeckBench.

First slide of Claude Fable 5.1's deck for Hearthside Coffee
Claude Fable 5.1 · deck rating 1,261 · needs work
First slide of Muse Spark 1.3's deck for Hearthside Coffee
Muse Spark 1.3 · deck rating 817 · needs work

Every slide, side by side →

From our research

Every shared result

Each model's best result on every measure both have, grouped by benchmark, with the setting that produced it. Under each measure, every model that benchmark has measured is a grey tick, best to the right, so you can see where the two sit in the field. The headline measures of each benchmark come first; the rest are under "Show all". Green marks the better value; where a source reports ranges that overlap, neither is marked.

MeasureClaude Fable 5.1Muse Spark 1.3
CatalogBench measured by Spring Prompt · 3 results
Cost per productUS dollars, lower is better 37 · 22 of 37 measured $0.0991 $0.0164
Reliably publish-ready% of products, higher is better 26 · 18 of 36 measured 26.8% 48.2%
Reliably publish-ready · sales brief% of products, higher is better 25 · 20 of 36 measured 5.4% 10.7%
DeckBench measured by Spring Prompt · 3 results
Cost per deckUS dollars, lower is better 35 · 14 of 37 measured $0.81 $0.0463
Deck ratingrating, higher is betterThe rating weighs each win by how strong the other deck was; the win rate counts wins against whichever decks a model happened to meet, so the two can disagree. 11 · 28 of 37 measured 1,261 817
Accurate decks% of tasks, higher is better 11 · 27 of 37 measured 66.7% 16.7%
ROASBench measured by Spring Prompt · 3 results
Overall scorescore out of 100, higher is better 12 · 28 of 37 measured 51.5 34.1
Audience scorescore out of 100, higher is better 9 · 29 of 37 measured 67.7 51.6
Business scorescore out of 100, higher is better 11 · 28 of 37 measured 52.3 25.8
APEX-Agents reported by APEX-Agents · 1 result
Tasks passed% of tasks, higher is better 3 · 12 of 37 measured 68.6% 57.8%
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Overall · TextArena rating, higher is better 6 · 8 of 387 measured 1,501max reasoning 1,494max reasoning
Overall · Text factualityArena rating, higher is better 3 · 13 of 163 measured 1,500max reasoning 1,480max reasoning
Business, management and finance · TextArena rating, higher is better 17 · 4 of 380 measured 1,484max reasoning 1,505max reasoning
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 3 · 10 of 249 measured 53.4max reasoning 48.1max reasoning
GPQA Diamond% of questions, higher is better 8 · 5 of 232 measured 93.7%max reasoning 94.1%xhigh reasoning
Humanity's Last Exam% of questions, higher is better 2 · 13 of 250 measured 59.1%max reasoning 48.7%max reasoning
GDP.pdf reported by GDP.pdf · 1 result
Professional document tasks% of rubric, higher is better 4 · 5 of 32 measured 29.6%high reasoning 27.6%
Show all 82 results (65 more)
MeasureClaude Fable 5.1Muse Spark 1.3
CatalogBench measured by Spring Prompt · 28 results
Channel compliance% of products, higher is better 1 · 1 of 37 measured 100.0% 100.0%
Channel rules broken% of products, lower is better 1 · 1 of 37 measured 0.0% 0.0%
Channel rules broken · sales brief% of products, lower is better 1 · 1 of 37 measured 0.0% 0.0%
Claims to check% of products, lower is better 26 · 1 of 35 measured 0.6% 0.0%
Claims to check · sales brief% of products, lower is better 1 · 1 of 35 measured 0.0% 0.0%
Conflicts caught% of conflicts, higher is better 1 · 1 of 37 measured 100.0% 100.0%
Content quality% of checks, higher is better 16 · 14 of 37 measured 97.6% 97.7%
Content quality · sales brief% of checks, higher is better 20 · 12 of 37 measured 97.2% 98.2%
Decision accuracy% of decisions, higher is better 25 · 12 of 37 measured 95.9% 97.6%
Failed outputs% of products, lower is better 1 · 1 of 37 measured 0.0% 0.0%
Failed outputs · sales brief% of products, lower is better 1 · 1 of 37 measured 0.0% 0.0%
Field accuracy% of missing fields, higher is better 8 · 25 of 37 measured 94.7% 92.9%
Invented values% of filled values, lower is better 7 · 7 of 37 measured 0.7% 0.7%
Missing UK information% of products, lower is better 1 · 1 of 37 measured 0.0% 0.0%
Missing UK information · sales brief% of products, lower is better 1 · 1 of 37 measured 0.0% 0.0%
Not findable% of products, lower is better 5 · 25 of 37 measured 1.8% 6.0%
Not findable · sales brief% of products, lower is better 5 · 24 of 37 measured 1.8% 5.4%
Publish-ready listings% of products, higher is better 28 · 18 of 37 measured 46.4% 65.5%
Publish-ready listings · sales brief% of products, higher is better 29 · 18 of 37 measured 13.7% 34.5%
Time per productseconds, lower is better 24 · 33 of 37 measured 21.7 47.5
Unsupported claimsclaims per product, lower is better 32 · 17 of 37 measured 1.23 0.44
Unsupported claims% of products, lower is better 34 · 17 of 37 measured 48.8% 14.9%
Unsupported claims · sales brief% of products, lower is better 34 · 19 of 37 measured 83.9% 57.7%
Unsupported claims · sales briefclaims per product, lower is better 33 · 18 of 37 measured 3.86 1.50
Wrong attributes% of products, lower is better 4 · 19 of 37 measured 11.9% 16.7%
Wrong attributes · sales brief% of products, lower is better 11 · 6 of 37 measured 14.9% 14.3%
Wrong category or variant% of products, lower is better 23 · 32 of 37 measured 2.4% 4.2%
Wrong category or variant · sales brief% of products, lower is better 7 · 30 of 37 measured 1.8% 3.6%
DeckBench measured by Spring Prompt · 14 results
Caveat dropped% of tasks, lower is better 1 · 16 of 37 measured 0.0% 16.7%
Clean layout% of tasks, higher is better 18 · 31 of 37 measured 33.3% 0.0%
Design quality% of the maximum, higher is better 20 · 28 of 37 measured 66.6% 56.1%
Draft figure quoted% of tasks, lower is better 21 · 15 of 37 measured 33.3% 16.7%
Findings missing% of tasks, lower is better 1 · 1 of 37 measured 0.0% 0.0%
Head-to-head win rate% of comparisons, higher is betterThe rating weighs each win by how strong the other deck was; the win rate counts wins against whichever decks a model happened to meet, so the two can disagree. 8 · 27 of 37 measured 75.0% 42.4%
Layout defects% of tasks, lower is better 18 · 31 of 37 measured 66.7% 100.0%
Misleading metric used% of tasks, lower is better 1 · 1 of 37 measured 0.0% 0.0%
Presentable decks% of tasks, higher is better 9 · 17 of 37 measured 33.3% 0.0%
Recommendation late or wrong% of tasks, lower is better 1 · 31 of 37 measured 0.0% 66.7%
Slides needing work% of tasks, lower is better 20 · 20 of 37 measured 50.0% 50.0%
Time per deckseconds, lower is better 21 · 14 of 37 measured 158 89.4
Unsupported claims% of tasks, lower is better 1 · 17 of 37 measured 0.0% 16.7%
Unsupported numbers% of tasks, lower is better 1 · 1 of 37 measured 0.0% 0.0%
ROASBench measured by Spring Prompt · 6 results
Consistency scorescore out of 100, higher is better 14 · 23 of 37 measured 33.3 26.7
Contribution profitsimulated US dollars, higher is better 2 · 28 of 37 measured $965,271 $337,251
Cost of a runUS dollars, lower is better 37 · 17 of 37 measured $2.49 $0.18
Months over budgetmonths of 12, lower is better 1 · 1 of 37 measured 0 0
Planning scorescore out of 100, higher is better 27 · 16 of 37 measured 54.6 54.8
Return on ad spendprofit per $1 spent, higher is better 14 · 28 of 37 measured $0.64 $0.29
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 9 results
Confirmed task success · AgentIPS effect estimate, higher is better 1 · 18 of 48 measured 0.18max reasoning, arena agent 0.05max reasoning, arena agent
Creative writing · TextArena rating, higher is better 9 · 22 of 385 measured 1,482max reasoning 1,459max reasoning
Expert prompts · TextArena rating, higher is better 15 · 12 of 338 measured 1,521max reasoning 1,529max reasoning
Instruction following · TextArena rating, higher is better 6 · 13 of 387 measured 1,497max reasoning 1,486max reasoning
Overall · DocumentArena rating, higher is better 4 · 15 of 38 measured 1,493max reasoning 1,468max reasoning
Praise over complaint · AgentIPS effect estimate, higher is better 2 · 21 of 48 measured 0.32max reasoning, arena agent 0.03max reasoning, arena agent
Steerability · AgentIPS effect estimate, higher is better 6 · 12 of 48 measured 0.09max reasoning, arena agent 0.05max reasoning, arena agent
Tool grounding · AgentIPS effect estimate, higher is better 4 · 14 of 48 measured 0.00max reasoning, arena agent 0.00max reasoning, arena agent
Writing, literature and language · TextArena rating, higher is better 7 · 19 of 386 measured 1,491max reasoning 1,470max reasoning
Artificial Analysis reported by Artificial Analysis · 8 results
Terminal-Bench 2.1% of tasks, higher is better 1 · 11 of 131 measured 91.4%max reasoning 85.4%xhigh reasoning
Terminal-Bench 4.0% of tasks, higher is better 8 · 19 of 120 measured 55.1%xhigh reasoning 33.3%max reasoning
τ-bench banking% of tasks, higher is better 8 · 3 of 121 measured 47.2%max reasoning 50.5%max reasoning
Artificial Analysis Coding Indexindex score, higher is better 1 · 7 of 130 measured 81.6max reasoning 76.5xhigh reasoning
Long-context reasoning (AA-LCR)% of questions, higher is better 4 · 16 of 241 measured 85.3%max reasoning 83.0%max reasoning
Output speedtokens per second, higher is better 58 · 9 of 78 measured 72.5max reasoning 254xhigh reasoning
SciCode% of problems, higher is better 2 · 10 of 121 measured 63.1%max reasoning 59.7%xhigh reasoning
Time to first answer tokenseconds, lower is better 24 · 66 of 78 measured 0.92low reasoning 29.6xhigh reasoning

Quick answers

Is Claude Fable 5.1 better than Muse Spark 1.3?

Pick Claude Fable 5.1 if quality matters most: it scores higher (Spring Prompt overall 64 against 59). Pick Muse Spark 1.3 if cost matters more: it costs 6.0 times less per product listing on CatalogBench. Muse Spark 1.3 is also faster.

Which is cheaper, Claude Fable 5.1 or Muse Spark 1.3?

Muse Spark 1.3 is cheaper to run: $0.016 against $0.099 per product listing on CatalogBench (6.0 times less).

Which is faster, Claude Fable 5.1 or Muse Spark 1.3?

Muse Spark 1.3 writes about 254 tokens a second on its usual API, against 72 for Claude Fable 5.1 (Artificial Analysis).

Which is better for product listings, Claude Fable 5.1 or Muse Spark 1.3?

Muse Spark 1.3 ranks 20 of 36 models on sale for turning a sparse product feed and photos into listings that can go live, against 27 for Claude Fable 5.1.

Which is better for decks from an analysis, Claude Fable 5.1 or Muse Spark 1.3?

Claude Fable 5.1 ranks 11 of 36 models on sale for turning a finished analysis into a deck you could present as it is, against 28 for Muse Spark 1.3.

Which is better for marketing planning, Claude Fable 5.1 or Muse Spark 1.3?

Claude Fable 5.1 ranks 12 of 36 models on sale for planning a year of ad spend without overspending, against 28 for Muse Spark 1.3.

Which is better for agents and tool use, Claude Fable 5.1 or Muse Spark 1.3?

Muse Spark 1.3 ranks 13 of 159 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 19 for Claude Fable 5.1.

Which has the bigger context window, Claude Fable 5.1 or Muse Spark 1.3?

Muse Spark 1.3: 1,048,576 tokens, against 1,000,000 for Claude Fable 5.1.

Run Claude Fable 5.1 and Muse Spark 1.3 on your own prompt.

Benchmarks aren't your task. Test both side by side on your prompt, with every answer checked against your rules and the cost of each one.