Compare

DeepSeek R2 vs GPT-5 Mini

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

Verdict

GPT-5 Mini is ~1.1× cheaper on a 3:1 input:output blend. DeepSeek R2 is open-weights, the other is proprietary.

DeepSeek

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

Frontier math at open-weights pricingLong CoT traces

OpenAI

GPT-5’s smaller, cheaper sibling. Optimized for chat and lightweight agent tasks.

Very low latencyCheap enough for mass consumer apps

Pricing

Input / 1M$0.55$0.50

Output / 1M$2.19$2.00

Context128K256K

Max output64K16K

LicenseOpen weightsProprietary

Released2026-01-152026-01-21

Benchmarks

LMSYS ELO—1288.0

MMLU Pro87.878.5

HumanEval90.185.5

MATH94.179.8

GPQA71.2—

IFEval—85.4

Per-task cost

Summarize a 1-hour meeting transcript$0.0093$0.0085

Review a 500-line pull request$0.0088$0.0080

Answer a customer support ticket$0.0034$0.0031

Extract structured data from a resume$0.0049$0.0045

Debug a stack trace with context$0.0077$0.0070

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

Add up to four models, tweak tasks live.