Compare

GPT-5 Turbo vs Claude Sonnet 4.6

OpenAI vs Anthropic. Specs, benchmarks, and real per-task cost — all in one page.

Verdict

GPT-5 Turbo leads on LMSYS ELO (1398 vs 1368). Claude Sonnet 4.6 is ~1.8× cheaper on a 3:1 input:output blend.

OpenAI

OpenAI’s flagship unified model. Handles text, vision, and audio natively. The generalist benchmark champion.

Native multimodalStrong math and scienceHuge ecosystem

Anthropic

The workhorse. Near-Opus quality at 1/5 the cost. The default choice for production code and agent workloads.

Excellent cost/quality ratioFast inferenceGreat for Cursor/Claude Code

Pricing

Input / 1M$6.00$3.00

Output / 1M$24.00$15.00

Context400K500K

Max output32K64K

LicenseProprietaryProprietary

Released2026-01-212025-11-04

Benchmarks

LMSYS ELO1398.01368.0

MMLU Pro91.288.7

HumanEval93.092.1

SWE-Bench65.864.5

MATH91.585.0

GPQA68.159.4

MMMU78.874.1

IFEval88.990.5

Per-task cost

Summarize a 1-hour meeting transcript$0.102$0.052

Review a 500-line pull request$0.096$0.054

Answer a customer support ticket$0.037$0.021

Extract structured data from a resume$0.054$0.030

Debug a stack trace with context$0.084$0.048

Per-call cost using published token counts for each task. Real-world prompts vary.

Add up to four models, tweak tasks live.