DeepSeek: DeepSeek V3 vs Google: Gemini 2.5 Pro

DeepSeek: DeepSeek V3 is cheaper than Google: Gemini 2.5 Pro by about 91% on a blended (75% input / 25% output) basis. DeepSeek: DeepSeek V3 charges $0.280000 / $0.420000 per 1M input/output tokens; Google: Gemini 2.5 Pro charges $1.250000 / $10.000000.

FieldDeepSeek: DeepSeek V3Google: Gemini 2.5 Pro
Input $/Mtok$0.280000$1.250000
Output $/Mtok$0.420000$10.000000
Cache read $/Mtok$0.028000$0.125000
Cache write $/Mtok––
Input $/Mchar0.091836n/a
Output $/Mchar0.137755n/a
Context window1638401048576
Tokenizerdeepseekgemini

Price history

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: DeepSeek V3Google: Gemini 2.5 Pro
Coding agent$2.87$16.88
RAG chat$3.01$25.63
Summarisation (batch)$2.94$21.25
Classification$2.83$14.25
Chat assistant$3.36$47.50
Content generation$3.92$82.50

Cite this page

DeepSeek: DeepSeek V3 vs Google: Gemini 2.5 Pro: $0.280000/$0.420000 vs $1.250000/$10.000000 per 1M tokens, as of 2026-10-04 (Token Barometer, https://tokenbarometer.com/compare/deepseek_deepseek-chat-vs-google_gemini-2.5-pro).

DeepSeek: DeepSeek V3 page | Google: Gemini 2.5 Pro page | Open in calculator