Compare AI Models

Put any models head-to-head β€” benchmarks, context window, parameters, weights and pricing in one view. Pick up to four; the πŸ† marks the leader on each benchmark.

Estimate monthly cost in β‚Ή β†’
Claude Fable 5.2GPT-5.5Claude Opus 4.8
Type
Multimodal
Multimodal
Multimodal
Context
1000K tokens
1050K tokens
1000K tokens
Parameters
β€”
β€”
β€”
Weights
πŸ”’ Proprietary
πŸ”’ Proprietary
πŸ”’ Proprietary
Benchmarks
MMLU
β€”
92
β€”
GPQA Diamond
β€”
94
94
SWE-bench Verified
β€”
89
89
AIME (math)
β€”
100
β€”
Pricing
API claude-fable-5-2: $8 / 1M input, $40 / 1M output (Fable 5.1 was $10 / $50); cache reads $0.20, 5-minute cache writes $10, 1-hour $16; batch $4 / $20; fast mode $16 / $80. 1M context, 128K max output. On the Claude API, Amazon Bedrock, Google Cloud and Microsoft Foundry.
API ~$5 / 1M input, ~$30 / 1M output. ChatGPT Plus β‚Ή1,999/mo (incl. GST), Pro tier for heaviest use; not on Free.
API ~$5 / 1M input, ~$25 / 1M output. Claude Pro/Max (~$20/mo, no India INR plan β€” ~β‚Ή1,800+ after FX/tax).

Benchmark scores are widely-reported public figures (higher is better) for relative comparison β€” they vary by source; check the provider for official results. πŸ† = best among the models shown.