Compare

Mistral X-Large vs Llama 4 Behemoth

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

Verdict

Llama 4 Behemoth leads on LMSYS ELO (1342 vs 1318). Mistral X-Large is ~1.3× cheaper on a 3:1 input:output blend.

Mistral

European frontier model. EU-hosted inference, strong European language coverage, Apache-licensed weights.

EU data residencyMultilingual strengthOpen license

Meta

Meta’s open-weights flagship. 405B params, fully open license, runs on every major inference provider.

Best open-weights modelRuns anywhereTransparent training

Pricing

Input / 1M$2.00$2.50

Output / 1M$6.00$8.00

Context256K256K

Max output16K16K

LicenseOpen weightsOpen weights

Released2026-01-282025-12-10

Benchmarks

LMSYS ELO1318.01342.0

MMLU Pro83.587.1

HumanEval88.089.3

SWE-Bench—52.4

MATH78.281.5

GPQA—56.2

IFEval86.588.0

Per-task cost

Summarize a 1-hour meeting transcript$0.033$0.041

Review a 500-line pull request$0.028$0.036

Answer a customer support ticket$0.011$0.014

Extract structured data from a resume$0.016$0.021

Debug a stack trace with context$0.024$0.031

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

Add up to four models, tweak tasks live.