Compare

Llama 4 Scout vs o4

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

Verdict

Llama 4 Scout is ~56.8× cheaper on a 3:1 input:output blend. Llama 4 Scout is open-weights, the other is proprietary.

Meta

The 70B workhorse open-weight. Fits on a single H100, beloved by self-hosters.

Single-GPU deployableDirt cheap on aggregatorsStrong fine-tuning base

OpenAI

Deep-reasoning model. Spends more tokens thinking to crush math, science, and hard coding problems.

State-of-the-art math and scienceExtremely high-ceiling reasoning

Pricing

Input / 1M$0.35$15.00

Output / 1M$0.80$60.00

Context128K200K

Max output8K100K

LicenseOpen weightsProprietary

Released2025-12-102026-02-18

Benchmarks

LMSYS ELO1265.0—

MMLU Pro77.293.8

HumanEval82.492.3

SWE-Bench—68.0

MATH72.596.2

GPQA—82.5

IFEval84.1—

Per-task cost

Summarize a 1-hour meeting transcript$0.0057$0.255

Review a 500-line pull request$0.0044$0.240

Answer a customer support ticket$0.0017$0.093

Extract structured data from a resume$0.0026$0.135

Debug a stack trace with context$0.0037$0.210

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

Add up to four models, tweak tasks live.