Compare

DeepSeek R2 vs Mistral X-Large

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

Verdict

DeepSeek R2 is ~3.1× cheaper on a 3:1 input:output blend.

DeepSeek

DeepSeek’s reasoning variant. Competes with o4 on math at a fraction of the cost.

Frontier math at open-weights pricingLong CoT traces

Mistral

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

EU data residencyMultilingual strengthOpen license

Pricing

Input / 1M$0.55$2.00

Output / 1M$2.19$6.00

Context128K256K

Max output64K16K

LicenseOpen weightsOpen weights

Released2026-01-152026-01-28

Benchmarks

LMSYS ELO—1318.0

MMLU Pro87.883.5

HumanEval90.188.0

MATH94.178.2

GPQA71.2—

IFEval—86.5

Per-task cost

Summarize a 1-hour meeting transcript$0.0093$0.033

Review a 500-line pull request$0.0088$0.028

Answer a customer support ticket$0.0034$0.011

Extract structured data from a resume$0.0049$0.016

Debug a stack trace with context$0.0077$0.024

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

Add up to four models, tweak tasks live.