Compare · ModelsLive · 3 picked · head to head
Claude Fable 5 vs GPT-5.6 Sol vs Kimi K3
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Claude Fable 5 wins on 16/24 benchmarks
Claude Fable 5 wins 16 of 24 shared benchmarks. Leads in agentic · reasoning · coding.
Category leads
agentic·Claude Fable 5reasoning·Claude Fable 5knowledge·GPT-5.6 Solcoding·Claude Fable 5general·Claude Fable 5math·Claude Fable 5speed·Claude Fable 5
Hype vs Reality
Attention vs performance
Claude Fable 5
#27 by perf·#12 by attention
GPT-5.6 Sol
#30 by perf·#4 by attention
Kimi K3
#51 by perf·#17 by attention
Best value
GPT-5.6 Sol
1.4x better value than Kimi K3
Claude Fable 5
2.3 pts/$
$30.00/M
GPT-5.6 Sol
11.1 pts/$
$6.00/M
Kimi K3
8.1 pts/$
$7.33/M
Vendor risk
Who is behind the model
Anthropic
$965.0B·Tier 1
OpenAI
$840.0B·Tier 1
Moonshot AI
$18.0B·Tier 1
Head to head
24 benchmarks · 3 models
Claude Fable 5GPT-5.6 SolKimi K3
APEX-Agents
Claude Fable 5 leads by +12.2
APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments.
Claude Fable 5
63.6
GPT-5.6 Sol
51.4
Kimi K3
50.6
ARC-AGI
Claude Fable 5 leads by +1.0
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
Claude Fable 5
98.5
GPT-5.6 Sol
97.5
Kimi K3
94.5
ARC-AGI-2
GPT-5.6 Sol leads by +3.3
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
89.2
GPT-5.6 Sol
92.5
Kimi K3
60.4
Chess Puzzles
GPT-5.6 Sol leads by +24.2
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
Claude Fable 5
37.9
GPT-5.6 Sol
62.1
Kimi K3
35.8
Deepswe
GPT-5.6 Sol leads by +2.8
Claude Fable 5
69.9
GPT-5.6 Sol
72.7
Kimi K3
68.5
Dtbench
Claude Fable 5 leads by +4.0
Claude Fable 5
97.3
GPT-5.6 Sol
93.3
Kimi K3
85.3
Frontiercode
Claude Fable 5 leads by +6.0
Claude Fable 5
53.5
GPT-5.6 Sol
47.5
Kimi K3
44.2
FrontierMath-Tier-4-v2-Private
Claude Fable 5 leads by +7.3
Claude Fable 5
90.2
GPT-5.6 Sol
82.9
Kimi K3
39.0
FrontierMath-Tiers-1-3-v2-Private
GPT-5.6 Sol leads by +2.1
Claude Fable 5
87.0
GPT-5.6 Sol
89.1
Kimi K3
72.2
Frontierswe
Claude Fable 5 leads by +14.8
Claude Fable 5
47.0
GPT-5.6 Sol
32.2
Kimi K3
25.9
Furniture Assembly
GPT-5.6 Sol leads by +29.8
Claude Fable 5
8.3
GPT-5.6 Sol
38.1
Kimi K3
6.0
GPQA diamond
GPT-5.6 Sol leads by +0.5
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
81.1
GPT-5.6 Sol
91.3
Kimi K3
90.8
Lmca
Claude Fable 5 leads by +2.3
Claude Fable 5
71.9
GPT-5.6 Sol
69.7
Kimi K3
62.0
Mystery Game Puzzles
GPT-5.6 Sol leads by +6.6
Claude Fable 5
47.1
GPT-5.6 Sol
53.7
Kimi K3
18.5
OTIS Mock AIME 2024-2025
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Claude Fable 5
100.0
GPT-5.6 Sol
100.0
Kimi K3
97.2
PostTrainBench
Claude Fable 5 leads by +5.6
Claude Fable 5
41.8
GPT-5.6 Sol
36.2
Kimi K3
32.0
Proofbench
Claude Fable 5 leads by +8.0
Claude Fable 5
95.0
GPT-5.6 Sol
83.0
Kimi K3
87.0
SimpleBench
Claude Fable 5 leads by +12.2
SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking.
Claude Fable 5
78.3
GPT-5.6 Sol
66.0
Kimi K3
52.8
SimpleQA Verified
Claude Fable 5 leads by +1.0
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
Claude Fable 5
70.7
GPT-5.6 Sol
69.7
Kimi K3
50.6
Surface Evolver Bench
Claude Fable 5
95.0
GPT-5.6 Sol
93.1
Kimi K3
95.0
WeirdML
Claude Fable 5 leads by +2.5
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
Claude Fable 5
91.9
GPT-5.6 Sol
89.4
Kimi K3
82.6
Artificial Analysis · Quality Index
Claude Fable 5 leads by +16.3
Claude Fable 5
59.9
Kimi K3
43.6
Ebr Bench
GPT-5.6 Sol leads by +5.2
Claude Fable 5
39.5
GPT-5.6 Sol
44.8
Mirrorcode
Claude Fable 5 leads by +43.9
Claude Fable 5
63.9
GPT-5.6 Sol
20.0
Full benchmark table
| Benchmark | Claude Fable 5 | GPT-5.6 Sol | Kimi K3 |
|---|---|---|---|
APEX-Agents APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments. | 63.6 | 51.4 | 50.6 |
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 98.5 | 97.5 | 94.5 |
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. | 89.2 | 92.5 | 60.4 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 37.9 | 62.1 | 35.8 |
Deepswe | 69.9 | 72.7 | 68.5 |
Dtbench | 97.3 | 93.3 | 85.3 |
Frontiercode | 53.5 | 47.5 | 44.2 |
FrontierMath-Tier-4-v2-Private | 90.2 | 82.9 | 39.0 |
FrontierMath-Tiers-1-3-v2-Private | 87.0 | 89.1 | 72.2 |
Frontierswe | 47.0 | 32.2 | 25.9 |
Furniture Assembly | 8.3 | 38.1 | 6.0 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 81.1 | 91.3 | 90.8 |
Lmca | 71.9 | 69.7 | 62.0 |
Mystery Game Puzzles | 47.1 | 53.7 | 18.5 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 100.0 | 100.0 | 97.2 |
PostTrainBench | 41.8 | 36.2 | 32.0 |
Proofbench | 95.0 | 83.0 | 87.0 |
SimpleBench SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking. | 78.3 | 66.0 | 52.8 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 70.7 | 69.7 | 50.6 |
Surface Evolver Bench | 95.0 | 93.1 | 95.0 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 91.9 | 89.4 | 82.6 |
Artificial Analysis · Quality Index | 59.9 | — | 43.6 |
Ebr Bench | 39.5 | 44.8 | — |
Mirrorcode | 63.9 | 20.0 | — |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $10.00 | $50.00 | 1.0M tokens (~500 books) | $200.00 | |
| $2.00 | $10.00 | 1.1M tokens (~525 books) | $40.00 | |
| $0.67 | $14.00 | 1.0M tokens (~524 books) | $40.03 |