Gemini 2.5 Flash token cost & pricing
Gemini 2.5 Flash is the price/performance workhorse of Google's lineup: $0.30 per million input, $2.50 per million output, and the full 1M context window. It is one of the few models that is simultaneously cheap and long-context capable, which makes it a common backbone for RAG and document pipelines.
Gemini 2.5 Flash specifications
| Vendor | |
|---|---|
| Input price | $0.3 / 1M tokens |
| Output price | $2.5 / 1M tokens |
| Cached input price | $0.075 / 1M tokens (75% cheaper) |
| Context window | 1,048,576 tokens |
| Max output | 65,536 tokens |
| Open weights | No |
| Price rank | 7 of 17 (cheapest first) |
| Best for | budget, long-context |
What Gemini 2.5 Flash costs in practice
Four common workloads, priced at list rates. Monthly figures assume 30 days.
| Workload | Per request | Requests/day | Per month |
|---|---|---|---|
| Chatbot · 1K in / 500 out | $0.0015 | 10,000 | $465.00 |
| RAG Q&A · 8K in / 400 out | $0.0034 | 5,000 | $510.00 |
| Doc summary · 30K in / 1K out | $0.01 | 1,000 | $345.00 |
| Code agent · 60K in / 3K out | $0.03 | 500 | $382.50 |
How many requests does $100 buy?
Assuming a 1,000-token prompt and a 500-token response per request.
How Gemini 2.5 Flash pricing actually works
About a quarter of Gemini 2.5 Pro's input price for the same context window. Cached input at $0.075 per million is 75% off. The 8.3x output ratio is the main thing to watch — a verbose answer erodes the savings quickly.
When to use Gemini 2.5 Flash
Ideal for high-volume document Q&A, classification over long inputs, and as the retrieval-side model in a two-stage pipeline where a stronger model handles final synthesis. Also a solid default for chat when latency and cost both matter.
How to cut your Gemini 2.5 Flash bill
Constrain output length explicitly in your prompt or via max tokens — at an 8x output ratio this is the highest-leverage change. Combine with caching for the static instruction prefix on repeated long-document queries.
Other Google models
Cheaper alternatives to Gemini 2.5 Flash
- DeepSeek R1 (reasoner)DeepSeek2% cheaper
- GPT-5 miniOpenAI20% cheaper
- Llama 4 MaverickMeta60% cheaper
- GPT-4o miniOpenAI73% cheaper
Run the numbers on your own prompt
Paste your actual text into the token counter to get an exact count instead of an estimate. Nothing is uploaded.
Open the calculator →