Claude Sonnet 4.6 vs GPT-4o: which is better?

Claude Sonnet 4.6 leads on combined benchmark scores (86.5 vs 85.2) across general knowledge, coding, and math, as of 2026-07-27.

Which is cheaper, Claude Sonnet 4.6 or GPT-4o?

GPT-4o is cheaper on a typical API workload, at roughly 1.4x lower cost than the alternative at published per-token rates.

Claude Sonnet 4.6 vs GPT-4o

Claude Sonnet 4.6 leads on benchmarks; GPT-4o is cheaper to run.

Updated 2026-07-27

Claude Sonnet 4.6

Anthropic · Closed

86.5

Combined index

Leads

GPT-4o

OpenAI · Closed

85.2

Combined index

Benchmarks

General knowledge (MMLU)

88.3

88.7

Coding

87.1

90.2

Math

84.2

76.6

API pricing

$3.0 Input / 1M $2.5

$15.0 Output / 1M $10.0

$135.0 Monthly, est. $100.0

Monthly estimate assumes 20M input + 5M output tokens. Prices marked ~ are representative hosted rates for open-weight models and vary by provider. Run your own numbers in the cost calculator.

Claude Sonnet 4.6 costs → GPT-4o costs → Full leaderboard → All comparisons →