Claude Fable 5.1 vs GPT-6 Astra
Updated October 20, 2026. 85 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Claude Fable 5.1 and GPT-6 Astra are level for coding, 89.1 against 88.7 on the Vibe Coding Index, so price and context decide.
Pick GPT-6 Astra when the context window is larger, 1.05M against 1M tokens, or when what matters is LMArena WebDev or Terminal-Bench 4 (Vals AI), where it leads. Pick Claude Fable 5.1 when what matters is LiveBench coding or AA Coding Index, where it leads.
- Shared results
- 85
- Claude Fable 5.1 only
- 39
- GPT-6 Astra only
- 7
- Like-for-like categories
- 8 of 9
Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.
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, four public coding benchmarks
No call
Level: 89.1 against 88.7, inside 2 points.
2 benchmarks behind each score
Intelligence
AA Intelligence Index
Claude Fable 5.1
Claude Fable 5.1 leads, 53.4 against 52.7.
Same board, both measured
Agentic work
AA Agentic Index
Claude Fable 5.1
Claude Fable 5.1 leads, 57.9 against 51.0.
Same board, both measured
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
GPT-6 Astra
GPT-6 Astra leads, 1338 against 1313.
Same board, both measured
General
Vals Index
No call
Only GPT-6 Astra has a Vals Index result, 65.0%.
Directional only
Long documents
Listed context window
GPT-6 Astra
GPT-6 Astra takes 1.05M tokens against 1M.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
No call
Claude Fable 5.1 $0.035, GPT-6 Astra $0.035 per turn.
List prices
At a glance
- Coding index
- Claude Fable 5.1 89.1 from 2 benchmarks, GPT-6 Astra 88.7 from 2 benchmarks.
- Widest category
- Agentic work, 14.2 points apart on average over 22 shared benchmarks.
- Widest single gap
- Vending-Bench 2: Claude Fable 5.1 $5,422 against GPT-6 Astra $15,515, 63.1 points apart on a 0 to 100 scale.
- Output price
- Both cost $50.00 per million output tokens.
- Context window
- Claude Fable 5.1 takes 1M tokens, GPT-6 Astra 1.05M.
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 | GPT-6 Astra | Ties |
|---|---|---|---|---|
| Coding | 10 | 12 | ||
| Intelligence | 1 | 1 | ||
| Agentic work | 8 | 14 | ||
| Reasoning | 3 | 9 | ||
| Math | 1 | 2 | 1 tie | |
| Knowledge | 1 | 1 | ||
| Web design | 2 | 7 | ||
| General | 10 | 2 | ||
| All shared boards | 36 | 48 | 1 tie |
CodingLike-for-likeFour of these boards feed the Vibe Coding Index1012
| Benchmark | Claude Fable 5.1 | GPT-6 Astra | Leader |
|---|---|---|---|
| LiveBench coding | 97.7% | 93.7% | Claude Fable 5.1 |
| LMArena WebDev | 1743 | 1774 | GPT-6 Astra |
| AA Coding Index | 81.6 | 76.9 | Claude Fable 5.1 |
| LiveCodeBench (Vals AI) | 90.5% | Not measured | Only Claude Fable 5.1 measured |
| Terminal-Bench 4 (Vals AI) | 58.1% | 59.6% | GPT-6 Astra |
| Vibe Code Bench (Vals AI) | 90.3% | 89.6% | Claude Fable 5.1 |
| Code migration (Vals AI) | 54.6% | 67.7% | GPT-6 Astra |
| ProgramBench (Vals AI) | 7.0% | 5.5% | Claude Fable 5.1 |
| IOI (Vals AI) | 90.8% | 100.0% | GPT-6 Astra |
| Terminal-Bench 2.1 (Vals AI) | 85.0% | 87.3% | GPT-6 Astra |
| SWE-bench Multimodal (Vals AI) | 61.3% | 43.3% | Claude Fable 5.1 |
| Vibe Code Bench 1-100 (Vals AI) | 28.0% | 27.6% | Claude Fable 5.1 |
| SkillsBench (Vals AI) | 61.5% | Not measured | Only Claude Fable 5.1 measured |
| CyberBench (Vals AI) | 70.4% | 41.1% | Claude Fable 5.1 |
| SciCode | 63.1% | 56.5% | Claude Fable 5.1 |
| DeepSWE | Not measured | 74.1% | Only GPT-6 Astra measured |
| FrontierCode | 50.9% | 53.3% | GPT-6 Astra |
| FrontierSWE | 56.3% | 65.5% | GPT-6 Astra |
| GSO-Bench | 88.2% | 79.4% | Claude Fable 5.1 |
| WeirdML | 92.9% | 93.3% | GPT-6 Astra |
| MirrorCode | 73.3% | 46.7% | Claude Fable 5.1 |
| ALE-Bench | 2143.2 | 2951.3 | GPT-6 Astra |
| LMArena Text: coding | 1531 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: software and IT | 1529 | Not measured | Only Claude Fable 5.1 measured |
| LMArena WebDev: React | 1792 | 1798 | GPT-6 Astra |
| LMArena WebDev: HTML | 1684 | 1775 | GPT-6 Astra |
| LMArena WebDev: from an image | 1722 | 1726 | GPT-6 Astra |
| Neither model has a result on the other 11 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index11
| Benchmark | Claude Fable 5.1 | GPT-6 Astra | Leader |
|---|---|---|---|
| AA Intelligence Index | 53.4 | 52.7 | Claude Fable 5.1 |
| Epoch Capabilities Index | 164.7 | 166.4 | GPT-6 Astra |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index814
| Benchmark | Claude Fable 5.1 | GPT-6 Astra | Leader |
|---|---|---|---|
| AA Agentic Index | 57.9 | 51.0 | Claude Fable 5.1 |
| Design Arena: agents building web apps | 1318 | 1373 | GPT-6 Astra |
| Design Arena: agents building full-stack apps | 1326 | 1357 | GPT-6 Astra |
| Design Arena: agents building mobile apps | 1266 | 1308 | GPT-6 Astra |
| Design Arena: agents building Godot games | 1327 | 1348 | GPT-6 Astra |
| Design Arena: PowerPoint slides | 1328 | 1245 | Claude Fable 5.1 |
| LiveBench research | 82.0% | 85.2% | GPT-6 Astra |
| LiveBench browser | 100.0% | 84.0% | Claude Fable 5.1 |
| LiveBench long agentic | 80.8% | 82.4% | GPT-6 Astra |
| SREBench (Vals AI) | 22.9% | 56.9% | GPT-6 Astra |
| Finance Agent (Vals AI) | 58.9% | 53.5% | Claude Fable 5.1 |
| Tax Agent Bench (Vals AI) | 49.2% | 20.7% | Claude Fable 5.1 |
| Legal Research Bench (Vals AI) | 55.3% | 39.4% | Claude Fable 5.1 |
| Harvey's Legal Agent Benchmark (Vals AI) | Not measured | 7.1% | Only GPT-6 Astra measured |
| Excel Modeling Benchmark (Vals AI) | 76.7% | 71.7% | Claude Fable 5.1 |
| BioMysteryBench (Vals AI) | Not measured | 79.3% | Only GPT-6 Astra measured |
| Terminal-Bench-Science (Vals AI) | 40.0% | 62.9% | GPT-6 Astra |
| Vals Science Index | Not measured | 65.1% | Only GPT-6 Astra measured |
| Time Horizon Index (Vals AI) | 63.3% | 90.5% | GPT-6 Astra |
| CUA-bench (Vals AI) | 13.2% | 19.2% | GPT-6 Astra |
| Robo-bench (Vals AI) | 5.8% | 23.3% | GPT-6 Astra |
| APEX-Agents | 68.6% | 64.7% | Claude Fable 5.1 |
| Remote Labor Index | 17.9% | 20.8% | GPT-6 Astra |
| BALROG | Not measured | 68.3% | Only GPT-6 Astra measured |
| SpireClimb | 2.0% | 25.7% | GPT-6 Astra |
| Vending-Bench 2 | $5,422 | $15,515 | GPT-6 Astra |
| LMArena Search | 1246 | Not measured | Only Claude Fable 5.1 measured |
| Neither model has a result on the other 12 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index39
| Benchmark | Claude Fable 5.1 | GPT-6 Astra | Leader |
|---|---|---|---|
| GPQA Diamond | 93.4% | Not measured | Only Claude Fable 5.1 measured |
| LiveBench reasoning | 91.7% | 92.7% | GPT-6 Astra |
| MysteryMechanism (Vals AI) | 47.8% | 53.1% | GPT-6 Astra |
| MMMU Pro (Vals AI) | 90.6% | Not measured | Only Claude Fable 5.1 measured |
| GPQA Diamond (Artificial Analysis run) | 93.7% | 96.3% | GPT-6 Astra |
| Humanity's Last Exam (Artificial Analysis run) | 59.1% | 54.7% | Claude Fable 5.1 |
| GPQA Diamond (Epoch AI run) | Not measured | 95.8% | Only GPT-6 Astra measured |
| ARC-AGI | 97.5% | 98.5% | GPT-6 Astra |
| ARC-AGI-2 | 90.0% | 95.0% | GPT-6 Astra |
| Humanity's Last Exam | 46.5% | 54.8% | GPT-6 Astra |
| Chess puzzles | 48.0% | 74.0% | GPT-6 Astra |
| Mystery game puzzles | 58.0% | 84.0% | GPT-6 Astra |
| EBR-bench | 57.1% | 76.2% | GPT-6 Astra |
| LMArena Text: hard prompts | 1526 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: expert | 1540 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: hard prompts in English | 1521 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Vision: diagrams | 1343 | 1297 | Claude Fable 5.1 |
| LMArena Vision: homework | 1345 | 1337 | Claude Fable 5.1 |
| Neither model has a result on the other 3 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index12
| Benchmark | Claude Fable 5.1 | GPT-6 Astra | Leader |
|---|---|---|---|
| OTIS Mock AIME | 100.0% | 100.0% | Tie |
| FrontierMath, tiers 1 to 3 | 90.2% | 93.7% | GPT-6 Astra |
| FrontierMath, tier 4 | 87.8% | 97.6% | GPT-6 Astra |
| ProofBench | 100.0% | 99.0% | Claude Fable 5.1 |
| LMArena Text: math | 1528 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: mathematical work | 1537 | 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 index11
| Benchmark | Claude Fable 5.1 | GPT-6 Astra | 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% | 48.5% | Claude Fable 5.1 |
| SimpleQA Verified | 70.8% | 75.6% | GPT-6 Astra |
| LMArena Text: science | 1532 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: medicine and healthcare | 1530 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: legal and government | 1511 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: business and finance | 1493 | Not measured | Only Claude Fable 5.1 measured |
| Neither model has a result on the other 7 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index27
| Benchmark | Claude Fable 5.1 | GPT-6 Astra | Leader |
|---|---|---|---|
| Design Arena: websites | 1313 | 1338 | GPT-6 Astra |
| Design Arena: UI components | 1339 | 1400 | GPT-6 Astra |
| Design Arena: game dev | 1381 | 1474 | GPT-6 Astra |
| Design Arena: data viz | 1344 | 1333 | Claude Fable 5.1 |
| Design Arena: 3D | 1393 | 1505 | GPT-6 Astra |
| Design Arena: all code categories | 1336 | 1379 | GPT-6 Astra |
| Design Arena: SVG | 1300 | 1431 | GPT-6 Astra |
| Design Arena: ASCII art | 1323 | 1356 | GPT-6 Astra |
| LiveBench design | 83.4% | 82.0% | Claude Fable 5.1 |
GeneralLike-for-likePublic boards, shown as published, kept out of the index102
| Benchmark | Claude Fable 5.1 | GPT-6 Astra | Leader |
|---|---|---|---|
| Vals Index | Not measured | 65.0% | Only GPT-6 Astra measured |
| LiveBench data analysis | 80.3% | 83.0% | GPT-6 Astra |
| Public Benefits Bench (Vals AI) | 74.9% | Not measured | Only Claude Fable 5.1 measured |
| SAGE (Vals AI) | 48.5% | 46.4% | Claude Fable 5.1 |
| MedScribe (Vals AI) | 91.3% | 87.9% | Claude Fable 5.1 |
| Long context reasoning (AA-LCR) | 85.3% | 80.7% | Claude Fable 5.1 |
| DTBench | 97.6% | 97.3% | Claude Fable 5.1 |
| LMCA | 65.5% | 64.4% | Claude Fable 5.1 |
| Furniture assembly | 70.0% | 80.0% | GPT-6 Astra |
| LMArena Text | 1512 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: instruction following | 1517 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: creative writing | 1508 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Vision | 1319 | 1280 | Claude Fable 5.1 |
| LMArena Text: English | 1505 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: non-English | 1507 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: Chinese | 1586 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: German | 1502 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: Japanese | 1560 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: Korean | 1521 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: French | 1527 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: Spanish | 1521 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: Russian | 1519 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: Polish | 1524 | 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 | 1523 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: writing and language | 1509 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Text: entertainment and media | 1495 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Vision: English | 1313 | 1280 | Claude Fable 5.1 |
| LMArena Vision: Chinese | 1391 | 1361 | Claude Fable 5.1 |
| LMArena Vision: reading text | 1327 | 1292 | Claude Fable 5.1 |
| LMArena Vision: humor | 1318 | Not measured | Only Claude Fable 5.1 measured |
| LMArena Vision: creative writing | 1330 | 1286 | Claude Fable 5.1 |
| Neither model has a result on the other 14 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Claude Fable 5.1 | GPT-6 Astra | 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
Same cost
Repository review
50K input + 3K output tokens
Same cost
Agent loop
200K input + 10K output tokens
Same cost
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 | OpenAI GPT-6 Astra |
|---|---|---|
| Maker | Anthropic | OpenAI |
| Weights | No public weights listed | No public weights listed |
| Released | September 1, 2026 | September 4, 2026 |
| Context window | 1M | 1.05M |
| Input $/Mtok | $10.00 | $10.00 |
| Output $/Mtok | $50.00 | $50.00 |
| Coding index | 89.1 | 88.7 |
| Independent measurements behind it | 2 | 2 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Claude Fable 5.1 better than GPT-6 Astra for coding?
Claude Fable 5.1 and GPT-6 Astra are level for coding, 89.1 against 88.7 on the Vibe Coding Index, so price and context decide. Pick GPT-6 Astra when the context window is larger, 1.05M against 1M tokens, or when what matters is LMArena WebDev or Terminal-Bench 4 (Vals AI), where it leads. Pick Claude Fable 5.1 when what matters is LiveBench coding or AA Coding Index, where it leads.
Which is cheaper, Claude Fable 5.1 or GPT-6 Astra?
Claude Fable 5.1 costs $10.00 per million input tokens and $50.00 per million output tokens; GPT-6 Astra costs $10.00 and $50.00. Cost per solved task is known for neither of them.
Which has the bigger context window, Claude Fable 5.1 or GPT-6 Astra?
GPT-6 Astra, with 1.05M tokens against 1M.
How many benchmarks is the Claude Fable 5.1 vs GPT-6 Astra comparison based on?
85 benchmarks have a result for both models as of October 20, 2026. Claude Fable 5.1's coding index rests on 2 independent benchmarks and GPT-6 Astra's on 2. Where only one model has a result, the page says so instead of naming a winner.