DeepSeek V3.2 (reasoner) vs DeepSeek V4 Pro

DeepSeek V3.2 (reasoner) is cheaper than DeepSeek V4 Pro by about 84% on a blended (75% input / 25% output) basis. DeepSeek V3.2 (reasoner) charges $0.28 / $0.42 per 1M input/output tokens; DeepSeek V4 Pro charges $1.32 / $3.96.

FieldDeepSeek V3.2 (reasoner)DeepSeek V4 Pro
Input $/Mtok$0.28$1.32
Output $/Mtok$0.42$3.96
Off-peak–off-peak −50% 04:00–06:00, 10:00–01:00 Mon–Fri; all day Sat–Sun UTC
Cache read $/Mtok$0.0280$0.0440
Cache write $/Mtok–free
Input $/Mchar0.09180.43
Output $/Mchar0.141.30
Context window1280001048576
Tokenizerdeepseekdeepseek
Confidence?aggregatorofficial
Sourcelitellmofficial-deepseek

Confidence: official = vendor pricing page · api = OpenRouter per-host · aggregator = community list, may lag · unreviewed = automatic extraction awaiting review. When sources disagree the most authoritative wins. Details: /methodology#confidence

Price history

No price changes recorded yet for either model.

Same workload, monthly cost

Illustrative monthly cost at 10M tokens/month, split per workload preset. Not a substitute for the calculator, which uses your own volume, cache and batch assumptions.

WorkloadDeepSeek V3.2 (reasoner)DeepSeek V4 Pro
Coding agent$2.87$14.52
RAG chat$3.01$17.16
Summarisation (batch)$2.94$15.84
Classification$2.83$13.73
Chat assistant$3.36$23.76
Content generation$3.92$34.32

Batch pricing not published for DeepSeek V3.2 (reasoner); standard rate used for the batch share of a workload.

Batch pricing not published for DeepSeek V4 Pro; standard rate used for the batch share of a workload.

Cite this page

DeepSeek V3.2 (reasoner) vs DeepSeek V4 Pro: $0.28/$0.42 vs $1.32/$3.96 per 1M tokens as of 2026-11-19. Token Barometer by Dejan Strbad, https://tokenbarometer.com, data CC BY 4.0. https://tokenbarometer.com/compare/deepseek_deepseek-reasoner-vs-deepseek_deepseek-v4-pro

DeepSeek V3.2 (reasoner) page | DeepSeek V4 Pro page | Open DeepSeek V3.2 (reasoner) in calculator | Open DeepSeek V4 Pro in calculator