Compare · ModelsLive · 3 picked · head to head
Claude Fable 5.2 vs Claude Opus 5.5 vs GPT-6 Astra
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Claude Fable 5.2 wins on 15/30 benchmarks
Claude Fable 5.2 wins 15 of 30 shared benchmarks. Leads in agentic · reasoning · general.
Category leads
agentic·Claude Fable 5.2reasoning·Claude Fable 5.2knowledge·GPT-6 Astrageneral·Claude Fable 5.2math·Claude Fable 5.2coding·Claude Opus 5.5speed·Claude Opus 5.5
Hype vs Reality
Attention vs performance
Claude Fable 5.2
#2 by perf·#5 by attention
Claude Opus 5.5
#11 by perf·#6 by attention
GPT-6 Astra
#9 by perf·#2 by attention
Best value
Claude Opus 5.5
1.7x better value than Claude Fable 5.2
Claude Fable 5.2
3.5 pts/$
$24.00/M
Claude Opus 5.5
6.2 pts/$
$12.00/M
GPT-6 Astra
2.5 pts/$
$30.00/M
Vendor risk
Who is behind the model
Anthropic
$965.0B·Tier 1
Anthropic
$965.0B·Tier 1
OpenAI
$840.0B·Tier 1
Head to head
30 benchmarks · 3 models
Claude Fable 5.2Claude Opus 5.5GPT-6 Astra
APEX-Agents
Claude Fable 5.2 leads by +6.6
APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments.
Claude Fable 5.2
80.1
Claude Opus 5.5
73.5
GPT-6 Astra
64.7
ARC-AGI
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
Claude Fable 5.2
98.5
Claude Opus 5.5
98.5
GPT-6 Astra
98.5
ARC-AGI-2
GPT-6 Astra leads by +1.3
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
Claude Fable 5.2
93.8
Claude Opus 5.5
93.3
GPT-6 Astra
95.0
Chess Puzzles
Claude Fable 5.2 leads by +3.2
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
Claude Fable 5.2
77.9
Claude Opus 5.5
65.3
GPT-6 Astra
74.8
Dtbench
Claude Fable 5.2 leads by +0.5
Claude Fable 5.2
98.7
Claude Opus 5.5
98.2
GPT-6 Astra
95.5
Ebr Bench
Claude Fable 5.2 leads by +14.3
Claude Fable 5.2
90.5
Claude Opus 5.5
71.4
GPT-6 Astra
76.2
FrontierMath-Tier-4-v2-Private
Claude Fable 5.2 leads by +2.4
Claude Fable 5.2
100.0
Claude Opus 5.5
95.0
GPT-6 Astra
97.6
FrontierMath-Tiers-1-3-v2-Private
Claude Fable 5.2 leads by +1.4
Claude Fable 5.2
100.0
Claude Opus 5.5
96.0
GPT-6 Astra
98.6
Frontierswe
Claude Fable 5.2 leads by +4.4
Claude Fable 5.2
69.9
Claude Opus 5.5
62.3
GPT-6 Astra
65.5
Furniture Assembly
Claude Fable 5.2 leads by +9.5
Claude Fable 5.2
85.7
Claude Opus 5.5
76.2
GPT-6 Astra
71.4
GPQA diamond
GPT-6 Astra leads by +2.6
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Claude Fable 5.2
95.7
Claude Opus 5.5
91.1
GPT-6 Astra
98.3
Lmca
Claude Fable 5.2 leads by +2.8
Claude Fable 5.2
83.1
Claude Opus 5.5
80.3
GPT-6 Astra
75.7
Mystery Game Puzzles
Claude Fable 5.2 leads by +6.6
Claude Fable 5.2
89.0
Claude Opus 5.5
68.0
GPT-6 Astra
82.4
OTIS Mock AIME 2024-2025
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Claude Fable 5.2
99.7
Claude Opus 5.5
100.0
GPT-6 Astra
100.0
Proofbench
Claude Fable 5.2
100.0
Claude Opus 5.5
100.0
GPT-6 Astra
99.0
Remote Labor Index
Claude Fable 5.2 leads by +20.0
Claude Fable 5.2
41.3
Claude Opus 5.5
21.3
GPT-6 Astra
20.8
SimpleBench
Claude Fable 5.2 leads by +7.8
SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking.
Claude Fable 5.2
93.9
Claude Opus 5.5
86.1
GPT-6 Astra
83.8
SimpleQA Verified
Claude Fable 5.2 leads by +18.4
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
Claude Fable 5.2
94.0
Claude Opus 5.5
72.2
GPT-6 Astra
75.6
SpireClimb
GPT-6 Astra leads by +14.4
Claude Fable 5.2
19.6
Claude Opus 5.5
5.8
GPT-6 Astra
34.0
Artificial Analysis · CritPt
Claude Opus 5.5
31.7
GPT-6 Astra
31.7
Artificial Analysis · GDPval
Claude Opus 5.5 leads by +14.7
Claude Opus 5.5
68.3
GPT-6 Astra
53.6
Artificial Analysis · Humanity's Last Exam
Claude Opus 5.5 leads by +6.7
Claude Opus 5.5
61.4
GPT-6 Astra
54.7
Artificial Analysis · Long Context Reasoning
Claude Opus 5.5 leads by +4.0
Claude Opus 5.5
84.7
GPT-6 Astra
80.7
Artificial Analysis · MMMU Pro
Claude Opus 5.5 leads by +0.8
Claude Opus 5.5
87.7
GPT-6 Astra
86.9
Artificial Analysis · Quality Index
Claude Opus 5.5 leads by +4.9
Claude Opus 5.5
57.6
GPT-6 Astra
52.7
Artificial Analysis · SciCode
Claude Opus 5.5 leads by +10.4
Claude Opus 5.5
66.9
GPT-6 Astra
56.5
Frontiercode
Claude Opus 5.5 leads by +1.4
Claude Opus 5.5
54.6
GPT-6 Astra
53.3
Mirrorcode
Claude Opus 5.5 leads by +35.6
Claude Opus 5.5
95.6
GPT-6 Astra
60.0
PostTrainBench
Claude Opus 5.5 leads by +1.9
Claude Opus 5.5
43.8
GPT-6 Astra
41.9
WeirdML v3
GPT-6 Astra leads by +4.7
Claude Opus 5.5
55.4
GPT-6 Astra
60.0
Full benchmark table
| Benchmark | Claude Fable 5.2 | Claude Opus 5.5 | GPT-6 Astra |
|---|---|---|---|
APEX-Agents APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments. | 80.1 | 73.5 | 64.7 |
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 98.5 | 98.5 | 98.5 |
ARC-AGI-2 ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data. | 93.8 | 93.3 | 95.0 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 77.9 | 65.3 | 74.8 |
Dtbench | 98.7 | 98.2 | 95.5 |
Ebr Bench | 90.5 | 71.4 | 76.2 |
FrontierMath-Tier-4-v2-Private | 100.0 | 95.0 | 97.6 |
FrontierMath-Tiers-1-3-v2-Private | 100.0 | 96.0 | 98.6 |
Frontierswe | 69.9 | 62.3 | 65.5 |
Furniture Assembly | 85.7 | 76.2 | 71.4 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 95.7 | 91.1 | 98.3 |
Lmca | 83.1 | 80.3 | 75.7 |
Mystery Game Puzzles | 89.0 | 68.0 | 82.4 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 99.7 | 100.0 | 100.0 |
Proofbench | 100.0 | 100.0 | 99.0 |
Remote Labor Index | 41.3 | 21.3 | 20.8 |
SimpleBench SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking. | 93.9 | 86.1 | 83.8 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 94.0 | 72.2 | 75.6 |
SpireClimb | 19.6 | 5.8 | 34.0 |
Artificial Analysis · CritPt | n/a | 31.7 | 31.7 |
Artificial Analysis · GDPval | n/a | 68.3 | 53.6 |
Artificial Analysis · Humanity's Last Exam | n/a | 61.4 | 54.7 |
Artificial Analysis · Long Context Reasoning | n/a | 84.7 | 80.7 |
Artificial Analysis · MMMU Pro | n/a | 87.7 | 86.9 |
Artificial Analysis · Quality Index | n/a | 57.6 | 52.7 |
Artificial Analysis · SciCode | n/a | 66.9 | 56.5 |
Frontiercode | n/a | 54.6 | 53.3 |
Mirrorcode | n/a | 95.6 | 60.0 |
PostTrainBench | n/a | 43.8 | 41.9 |
WeirdML v3 | n/a | 55.4 | 60.0 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $8.00 | $40.00 | 1.0M tokens (~500 books) | $160.00 | |
| $4.00 | $20.00 | 1.0M tokens (~500 books) | $80.00 | |
| $10.00 | $50.00 | 1.1M tokens (~525 books) | $200.00 |
People also compared