Compare

DeepSeek R2 vs Qwen 3 Max

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

Verdict

DeepSeek R2 is ~2.9× 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

Qwen

Alibaba’s flagship. Strongest non-English coverage of any open model; dominant in Asian markets.

MultilingualOpen weightsCompetitive coding

Pricing

Input / 1M$0.55$1.60

Output / 1M$2.19$6.40

Context128K128K

Max output64K16K

LicenseOpen weightsOpen weights

Released2026-01-152025-11-20

Benchmarks

LMSYS ELO—1295.0

MMLU Pro87.884.0

HumanEval90.187.2

MATH94.185.1

GPQA71.2—

MMMU—71.8

IFEval—85.5

Per-task cost

Summarize a 1-hour meeting transcript$0.0093$0.027

Review a 500-line pull request$0.0088$0.026

Answer a customer support ticket$0.0034$0.0099

Extract structured data from a resume$0.0049$0.014

Debug a stack trace with context$0.0077$0.022

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

Add up to four models, tweak tasks live.