Gemini 3.5 Flash vs Gemini 2.5 Pro: cost compared
Last verified
Gemini 2.5 Pro is 17% cheaper on input and -11% cheaper on output than Gemini 3.5 Flash. Whether that makes it the right choice depends entirely on whether it does your task well enough — this page gives you the cost side of that trade, precisely.
Rates side by side
| $ per million tokens | Gemini 3.5 Flash | Gemini 2.5 Pro |
|---|---|---|
| Input | $1.50 | $1.25 |
| Output | $9.00 | $10.00 |
| Cached input | $0.15 | $0.13 |
| Output : input ratio | 6.0× | 8.0× |
What the difference is worth
Monthly cost for the same workload on each, without caching:
| Workload | Gemini 3.5 Flash | Gemini 2.5 Pro | Difference |
|---|---|---|---|
| Support chatbot 1,000 conversations/month · 2,000 in + 500 out each | $7.50 | $7.50 | $0.0000 saved |
| RAG search 10,000 queries/month · 8,000 in + 400 out each | $156 | $140 | $16.00 saved |
| Bulk extraction 10,000 documents · 3,000 in + 300 out each | $72.00 | $67.50 | $4.50 saved |
| Coding agent 100 runs/month · 400,000 in + 20,000 out each | $78.00 | $70.00 | $8.00 saved |
Which to use
Gemini 2.5 Pro if the task is well-specified and mechanical — extraction, classification, formatting, routine transformation. The saving is real and compounds with volume.
Gemini 3.5 Flash if the task involves genuine reasoning, long-horizon planning, or work where a wrong answer is expensive to catch. Paying 1.2× more on input is trivial compared to the cost of shipping a bad result.
Both are Google models, so switching is usually a one-line change and the tokenizer is the same — which makes this a genuinely easy experiment to run. Test on your own workload before deciding.
Don't skip caching
Cached input costs $0.15 on Gemini 3.5 Flash and $0.13 on Gemini 2.5 Pro. If your prompts share a stable prefix, caching on the more expensive model can beat switching to the cheaper one outright — worth checking before you migrate anything.
Sources
Google rates verified 2026-08-05 (source). Google rates verified 2026-08-05 (source).