Compare

GPT-5 Turbo vs Grok 4

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

Verdict

GPT-5 Turbo leads on LMSYS ELO (1398 vs 1358). Grok 4 is ~1.4× 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

xAI

xAI’s latest. Real-time X search integration, strong on current events and meme-literate tasks.

Live web/X accessPersonalityFast inference on Colossus

Pricing

Input / 1M$6.00$5.00

Output / 1M$24.00$15.00

Context400K256K

Max output32K32K

LicenseProprietaryProprietary

Released2026-01-212026-03-01

Benchmarks

LMSYS ELO1398.01358.0

MMLU Pro91.286.2

HumanEval93.088.4

SWE-Bench65.8—

MATH91.586.5

GPQA68.1—

MMMU78.8—

IFEval88.985.0

Per-task cost

Summarize a 1-hour meeting transcript$0.102$0.082

Review a 500-line pull request$0.096$0.070

Answer a customer support ticket$0.037$0.027

Extract structured data from a resume$0.054$0.040

Debug a stack trace with context$0.084$0.060

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

Add up to four models, tweak tasks live.