LLM prices / Compare
API price comparison · checked Oct 23, 2026
Gemini 2.5 Pro vs Gemini 3.1 Pro Preview: API price
Short answerGemini 2.5 Pro is the cheaper one: 1,000 chatbot replies cost $5.88 against $7.80 on Gemini 3.1 Pro Preview, so Gemini 3.1 Pro Preview costs 1.33x as much. Across our 6 workloads Gemini 2.5 Pro is cheaper on 6 and Gemini 3.1 Pro Preview on 0. Standard prices: Gemini 2.5 Pro $1.25 / $10, Gemini 3.1 Pro Preview $2 / $12 per 1M input / output tokens, checked Oct 23, 2026.
Cost of the same work
| Workload (1,000 requests) | Gemini 2.5 Pro | Gemini 3.1 Pro Preview | Ratio |
|---|---|---|---|
| Chatbot replya 1,500-token prompt (1,000 of it a fixed system prompt) and a 400-token answer | $5.88 | $7.80 | 0.75x |
| Answer from your docs (RAG)6,000 tokens of retrieved context plus the question, a 500-token answer | $12.50 | $18.00 | 0.69x |
| Summarize a long documenta 20,000-token document in, an 800-token summary out | $33.00 | $49.60 | 0.67x |
| Coding agent step40,000 tokens of repo context and history (30,000 of it repeated from the last step), 2,000 out | $70.00 | $104 | 0.67x |
| Batch extraction2,000 tokens per record in, 300 tokens of JSON out, sent through the batch API where there is one | $2.75 | $3.80 | 0.72x |
| Long-document Q&Aa 150,000-token contract or codebase in one prompt, a 1,000-token answer | $198 | $312 | 0.63x |
Side by side
| Price | Gemini 2.5 Pro | Gemini 3.1 Pro Preview |
|---|---|---|
| Input / 1M | $1.25 | $2 |
| Output / 1M | $10 | $12 |
| Cache read / 1M | $0.125 | $0.2 |
| Batch input / output | $0.625 / $5 | $1 / $6 |
| Long-prompt tier | over 200,000 | over 200,000 |
| Provider | ||
| Status | listed | listed |
When each one is cheaper
- Gemini 2.5 Pro is cheaper or equal on both input and output, so it costs less on every workload at standard prices.
- With caching on a coding-agent step (30,000 of 40,000 input tokens cached): Gemini 2.5 Pro $36.25 vs Gemini 3.1 Pro Preview $50.00 per 1,000 steps (0.72x).
- This page compares price only. Which model answers better for your task is a separate question: test both on your own prompts.
Each model in full
Sources
- Google API pricing: https://ai.google.dev/gemini-api/docs/pricing, checked Oct 23, 2026.