Claude Fable 5.1 vs Claude Haiku 5.5
Claude Fable 5.1 leads Claude Haiku 5.5 on coding, 84.2 to 70.7 on the Vibe Coding Index, and Claude Haiku 5.5 costs 100.0x less per output token ($0.50 vs $50.00 per million). Updated October 22, 2026; 58 benchmarks have a result for both.
vs boards won of 58 shared
Claude Fable 5.1 or Claude Haiku 5.5: which to pick
Pick Claude Fable 5.1 for coding: it scores 84.2 on the Vibe Coding Index against 70.7 for Claude Haiku 5.5.
Pick Claude Haiku 5.5 for output tokens 100.0x cheaper ($0.50 vs $50.00 per million) and work measured by Vibe Code Bench (Vals AI) or CyberBench (Vals AI), where it leads.
Model AClaude Fable 5.1Anthropic84.2/1005 independent benchmarksNo cost per solved taskModel BClaude Haiku 5.5Anthropic70.7/1005 independent benchmarksNo cost per solved task
- Shared results
- 58
- Claude Fable 5.1 only
- 75
- Claude Haiku 5.5 only
- 7
- Like-for-like categories
- 8 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
Claude Fable 5.1
Claude Fable 5.1 scores 84.2 against 70.7.
5 benchmarks behind each score
Intelligence
AA Intelligence Index
Claude Fable 5.1
Claude Fable 5.1 leads, 53.4 against 43.4.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only Claude Fable 5.1 has a AA Agentic Index result, 57.9.
Directional only
Reasoning
GPQA Diamond
No call
Only Claude Fable 5.1 has a GPQA Diamond result, 93.4%.
Directional only
Knowledge
MMLU Pro
No call
Only Claude Fable 5.1 has a MMLU Pro result, 92.4%.
Directional only
Web design
Design Arena: websites
Claude Fable 5.1
Claude Fable 5.1 leads, 1312 against 1262.
Same board, both measured
General
Vals Index
No call
Only Claude Haiku 5.5 has a Vals Index result, 58.7%.
Directional only
Long documents
Listed context window
No call
Both take 1M tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Claude Haiku 5.5
Claude Fable 5.1 $0.035, Claude Haiku 5.5 $0.0003 per turn.
List prices
At a glance
- Coding index
- Claude Fable 5.1 84.2 from 5 benchmarks, Claude Haiku 5.5 70.7 from 5 benchmarks.
- Widest category
- Knowledge, 24.5 points apart on average over 2 shared benchmarks.
- Widest single gap
- SimpleQA Verified: Claude Fable 5.1 70.8% against Claude Haiku 5.5 23.8%, 47.0 points apart on a 0 to 100 scale.
- Output price
- Claude Fable 5.1 $50.00 against Claude Haiku 5.5 $0.50 per million output tokens, 100.0x apart.
- Context window
- Both take 1M tokens.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | Claude Fable 5.1 | Bars | Claude Haiku 5.5 | Ties |
|---|---|---|---|---|
| Coding | 14 | 2 | ||
| Intelligence | 2 | 0 | ||
| Agentic work | 11 | 0 | ||
| Reasoning | 7 | 1 | ||
| Math | 4 | 0 | ||
| Knowledge | 2 | 0 | ||
| Web design | 7 | 0 | ||
| General | 8 | 0 | ||
| All shared boards | 55 | 3 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index142
| Benchmark | Claude Fable 5.1 | Claude Haiku 5.5 | Leader |
|---|---|---|---|
| LiveBench coding | 97.7% | 91.4% | Claude Fable 5.1 |
| LMArena WebDev | 1743 | 1586 | Claude Fable 5.1 |
| AA Coding Index | 81.6 | Not measured | Only Claude Fable 5.1 measured |
| LiveCodeBench (Vals AI) | 90.5% | Not measured | Only Claude Fable 5.1 measured |
| Terminal-Bench 4 (Vals AI) | 58.1% | 36.4% | Claude Fable 5.1 |
| Vibe Code Bench (Vals AI) | 90.3% | 90.4% | Claude Haiku 5.5 |
| Code migration (Vals AI) | 54.6% | 44.6% | Claude Fable 5.1 |
| ProgramBench (Vals AI) | 7.0% | 2.5% | Claude Fable 5.1 |
| IOI (Vals AI) | 90.8% | 47.3% | Claude Fable 5.1 |
| Terminal-Bench 2.1 (Vals AI) | 85.0% | Not measured | Only Claude Fable 5.1 measured |
| SWE-bench Multimodal (Vals AI) | 61.3% | 44.0% | Claude Fable 5.1 |
| Vibe Code Bench 1-100 (Vals AI) | 28.0% | 20.4% | Claude Fable 5.1 |
| SkillsBench (Vals AI) | 61.5% | Not measured | Only Claude Fable 5.1 measured |
| CyberBench (Vals AI) | 70.4% | 75.5% | Claude Haiku 5.5 |
| SciCode | 63.1% | 55.0% | Claude Fable 5.1 |
| FrontierCode | 50.9% | 46.4% | Claude Fable 5.1 |
| FrontierSWE | 56.3% | 43.8% | Claude Fable 5.1 |
| GSO-Bench | 88.2% | Not measured | Only Claude Fable 5.1 measured |
| WeirdML | 92.9% | Not measured | Only Claude Fable 5.1 measured |
| MirrorCode | 73.3% | Not measured | Only Claude Fable 5.1 measured |
| ALE-Bench | 2143.2 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: coding | 1532 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: software and IT | 1530 | Not measured | Only Claude Fable 5.1 measured |
| LMArena WebDev: React | 1792 | 1605 | Claude Fable 5.1 |
| LMArena WebDev: HTML | 1684 | 1569 | Claude Fable 5.1 |
| LMArena WebDev: from an image | 1722 | Not measured | Only Claude Fable 5.1 measured |
| SciCode | Not measured | 9.2% | Only Claude Haiku 5.5 measured |
| SWE-Bench Pro V2 | 99.1% | 87.2% | Claude Fable 5.1 |
| SWE Atlas: Codebase QnA | 60.0% | Not measured | Only Claude Fable 5.1 measured |
| SWE Atlas: Refactoring | 56.7% | Not measured | Only Claude Fable 5.1 measured |
| SWE Atlas: Test Writing | 67.0% | Not measured | Only Claude Fable 5.1 measured |
| Neither model has a result on the other 12 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Claude Fable 5.1 | Claude Haiku 5.5 | Leader |
|---|---|---|---|
| AA Intelligence Index | 53.4 | 43.4 | Claude Fable 5.1 |
| Epoch Capabilities Index | 166.5 | 159.3 | Claude Fable 5.1 |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index110
| Benchmark | Claude Fable 5.1 | Claude Haiku 5.5 | Leader |
|---|---|---|---|
| AA Agentic Index | 57.9 | Not measured | Only Claude Fable 5.1 measured |
| Design Arena: agents building web apps | 1318 | Not measured | Only Claude Fable 5.1 measured |
| Design Arena: agents building full-stack apps | 1326 | Not measured | Only Claude Fable 5.1 measured |
| Design Arena: agents building mobile apps | 1267 | Not measured | Only Claude Fable 5.1 measured |
| Design Arena: agents building Godot games | 1327 | Not measured | Only Claude Fable 5.1 measured |
| Design Arena: PowerPoint slides | 1307 | Not measured | Only Claude Fable 5.1 measured |
| Design Arena: spreadsheets | 1394 | Not measured | Only Claude Fable 5.1 measured |
| LiveBench research | 82.0% | 81.9% | Claude Fable 5.1 |
| LiveBench browser | 100.0% | 80.0% | Claude Fable 5.1 |
| LiveBench long agentic | 80.8% | 57.6% | Claude Fable 5.1 |
| SREBench (Vals AI) | 22.9% | 5.7% | Claude Fable 5.1 |
| Finance Agent (Vals AI) | 58.9% | 54.1% | Claude Fable 5.1 |
| Tax Agent Bench (Vals AI) | 49.2% | 31.9% | Claude Fable 5.1 |
| Legal Research Bench (Vals AI) | 55.3% | 43.3% | Claude Fable 5.1 |
| Harvey's Legal Agent Benchmark (Vals AI) | Not measured | 4.2% | Only Claude Haiku 5.5 measured |
| Excel Modeling Benchmark (Vals AI) | 76.7% | 59.0% | Claude Fable 5.1 |
| BioMysteryBench (Vals AI) | Not measured | 70.4% | Only Claude Haiku 5.5 measured |
| Terminal-Bench-Science (Vals AI) | 40.0% | 10.0% | Claude Fable 5.1 |
| Vals Science Index | Not measured | 38.5% | Only Claude Haiku 5.5 measured |
| Time Horizon Index (Vals AI) | 63.3% | Not measured | Only Claude Fable 5.1 measured |
| CUA-bench (Vals AI) | 13.2% | Not measured | Only Claude Fable 5.1 measured |
| Robo-bench (Vals AI) | 5.8% | Not measured | Only Claude Fable 5.1 measured |
| APEX-Agents | 68.6% | 56.1% | Claude Fable 5.1 |
| Remote Labor Index | 17.9% | Not measured | Only Claude Fable 5.1 measured |
| SpireClimb | 2.0% | Not measured | Only Claude Fable 5.1 measured |
| Vending-Bench 2 | $5,422 | $3,602 | Claude Fable 5.1 |
| LMArena Search | 1246 | Not measured | Only Claude Fable 5.1 measured |
| MCP Atlas | 87.2% | Not measured | Only Claude Fable 5.1 measured |
| Neither model has a result on the other 14 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index71
| Benchmark | Claude Fable 5.1 | Claude Haiku 5.5 | Leader |
|---|---|---|---|
| GPQA Diamond | 93.4% | Not measured | Only Claude Fable 5.1 measured |
| LiveBench reasoning | 91.7% | 82.4% | Claude Fable 5.1 |
| MysteryMechanism (Vals AI) | 47.8% | 35.1% | Claude Fable 5.1 |
| MMMU Pro (Vals AI) | 90.6% | Not measured | Only Claude Fable 5.1 measured |
| GPQA Diamond (Artificial Analysis run) | 93.7% | Not measured | Only Claude Fable 5.1 measured |
| Humanity's Last Exam (Artificial Analysis run) | 59.1% | 44.4% | Claude Fable 5.1 |
| GPQA Diamond (Epoch AI run) | Not measured | 89.6% | Only Claude Haiku 5.5 measured |
| ARC-AGI | 97.5% | 89.0% | Claude Fable 5.1 |
| ARC-AGI-2 | 90.0% | 61.3% | Claude Fable 5.1 |
| Chess puzzles | 48.0% | 55.0% | Claude Haiku 5.5 |
| Mystery game puzzles | 58.0% | 30.0% | Claude Fable 5.1 |
| EBR-bench | 57.1% | 43.8% | Claude Fable 5.1 |
| LMArena Text: hard prompts | 1528 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: expert | 1544 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: hard prompts in English | 1522 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Vision: diagrams | 1343 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Vision: homework | 1345 | Not measured | Only Claude Fable 5.1 measured |
| Humanity's Last Exam (Scale run) | 51.3% | Not measured | Only Claude Fable 5.1 measured |
| Neither model has a result on the other 5 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index40
| Benchmark | Claude Fable 5.1 | Claude Haiku 5.5 | Leader |
|---|---|---|---|
| OTIS Mock AIME | 100.0% | 98.9% | Claude Fable 5.1 |
| FrontierMath, tiers 1 to 3 | 90.2% | 75.1% | Claude Fable 5.1 |
| FrontierMath, tier 4 | 87.8% | 46.3% | Claude Fable 5.1 |
| ProofBench | 100.0% | 89.0% | Claude Fable 5.1 |
| LMArena Text: math | 1531 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: mathematical work | 1541 | Not measured | Only Claude Fable 5.1 measured |
| MathArena | 78.1% | Not measured | Only Claude Fable 5.1 measured |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Claude Fable 5.1 | Claude Haiku 5.5 | Leader |
|---|---|---|---|
| MMLU Pro | 92.4% | Not measured | Only Claude Fable 5.1 measured |
| LegalBench (Vals AI) | 88.5% | Not measured | Only Claude Fable 5.1 measured |
| TaxEval (Vals AI) | 76.0% | Not measured | Only Claude Fable 5.1 measured |
| MortgageTax (Vals AI) | 70.8% | Not measured | Only Claude Fable 5.1 measured |
| MedCode (Vals AI) | 53.5% | 51.5% | Claude Fable 5.1 |
| SimpleQA Verified | 70.8% | 23.8% | Claude Fable 5.1 |
| LMArena Text: science | 1532 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: medicine and healthcare | 1531 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: legal and government | 1514 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: business and finance | 1493 | Not measured | Only Claude Fable 5.1 measured |
| PRBench Finance | 50.8% | Not measured | Only Claude Fable 5.1 measured |
| PRBench Legal | 51.6% | Not measured | Only Claude Fable 5.1 measured |
| Neither model has a result on the other 8 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index70
| Benchmark | Claude Fable 5.1 | Claude Haiku 5.5 | Leader |
|---|---|---|---|
| Design Arena: websites | 1312 | 1262 | Claude Fable 5.1 |
| Design Arena: UI components | 1336 | 1283 | Claude Fable 5.1 |
| Design Arena: game dev | 1381 | 1315 | Claude Fable 5.1 |
| Design Arena: data viz | 1345 | 1297 | Claude Fable 5.1 |
| Design Arena: 3D | 1392 | 1342 | Claude Fable 5.1 |
| Design Arena: all code categories | 1335 | 1278 | Claude Fable 5.1 |
| Design Arena: SVG | 1300 | Not measured | Only Claude Fable 5.1 measured |
| Design Arena: ASCII art | 1323 | Not measured | Only Claude Fable 5.1 measured |
| LiveBench design | 83.4% | 83.1% | Claude Fable 5.1 |
GeneralLike-for-likePublic boards, shown as published, kept out of the index80
| Benchmark | Claude Fable 5.1 | Claude Haiku 5.5 | Leader |
|---|---|---|---|
| Vals Index | Not measured | 58.7% | Only Claude Haiku 5.5 measured |
| LiveBench data analysis | 80.3% | 53.1% | Claude Fable 5.1 |
| Public Benefits Bench (Vals AI) | 74.9% | 63.2% | Claude Fable 5.1 |
| SAGE (Vals AI) | 48.5% | 45.8% | Claude Fable 5.1 |
| MedScribe (Vals AI) | 91.3% | 89.5% | Claude Fable 5.1 |
| Long context reasoning (AA-LCR) | 85.3% | 82.7% | Claude Fable 5.1 |
| DTBench | 97.6% | 95.0% | Claude Fable 5.1 |
| LMCA | 65.5% | 51.1% | Claude Fable 5.1 |
| Furniture assembly | 70.0% | 47.5% | Claude Fable 5.1 |
| LMArena Text | 1512 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: instruction following | 1518 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: creative writing | 1509 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Vision | 1319 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: English | 1506 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: non-English | 1508 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: Chinese | 1584 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: German | 1499 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: Japanese | 1562 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: Korean | 1517 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: French | 1523 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: Spanish | 1520 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: Russian | 1520 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: Polish | 1525 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: multi-turn | 1493 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: longer queries | 1524 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: writing and language | 1510 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: entertainment and media | 1497 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Vision: English | 1313 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Vision: Chinese | 1391 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Vision: reading text | 1327 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Vision: humor | 1318 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Vision: creative writing | 1330 | Not measured | Only Claude Fable 5.1 measured |
| MultiChallenge | Not measured | 67.7% | Only Claude Haiku 5.5 measured |
| Neither model has a result on the other 15 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Claude Fable 5.1 | Claude Haiku 5.5 | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Claude Fable 5.1$0.035Fits in one request
Claude Haiku 5.5$0.0003Fits in one request
Claude Haiku 5.5 costs less
Repository review
50K input + 3K output tokens
Claude Fable 5.1$0.650Fits in one request
Claude Haiku 5.5$0.0065Fits in one request
Claude Haiku 5.5 costs less
Agent loop
200K input + 10K output tokens
Claude Fable 5.1$2.50Fits in one request
Claude Haiku 5.5$0.025Fits in one request
Claude Haiku 5.5 costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | Anthropic Claude Fable 5.1 | Anthropic Claude Haiku 5.5 |
|---|---|---|
| Maker | Anthropic | Anthropic |
| Weights | No public weights listed | No public weights listed |
| Released | September 1, 2026 | October 7, 2026 |
| Context window | 1M | 1M |
| Input $/Mtok | $10.00 | $0.10 |
| Output $/Mtok | $50.00 | $0.50 |
| Coding index | 84.2 | 70.7 |
| Independent measurements behind it | 5 | 5 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Claude Fable 5.1 better than Claude Haiku 5.5 for coding?
Pick Claude Fable 5.1 for coding: it scores 84.2 on the Vibe Coding Index against 70.7 for Claude Haiku 5.5. Pick Claude Haiku 5.5 for output tokens 100.0x cheaper ($0.50 vs $50.00 per million) and work measured by Vibe Code Bench (Vals AI) or CyberBench (Vals AI), where it leads.
Which is cheaper, Claude Fable 5.1 or Claude Haiku 5.5?
Claude Fable 5.1 costs $10.00 per million input tokens and $50.00 per million output tokens; Claude Haiku 5.5 costs $0.10 and $0.50. Cost per solved task is known for neither of them.
Which has the bigger context window, Claude Fable 5.1 or Claude Haiku 5.5?
Neither: both take 1M tokens.
How many benchmarks is the Claude Fable 5.1 vs Claude Haiku 5.5 comparison based on?
58 benchmarks have a result for both models as of October 22, 2026. Claude Fable 5.1's coding index rests on 5 independent benchmarks and Claude Haiku 5.5's on 5. Where only one model has a result, the page says so instead of naming a winner.