Gemini 3 Flash
Gemini 3 Flash — Google's fastest capable model with 1M token context at $0.50/M input.
| Type | Price / 1M tokens | Notes |
|---|---|---|
| Input | $0.50 | Per 1M input tokens |
| Output | $3.00 | Per 1M output tokens |
| Cache read | $0.13 | Cached prefix read-back |
Gemini 3 Flash undercuts GPT-5 mini on price while offering a 1M token context window. The obvious choice for long-document pipelines that need to stay cheap.
Best for
- Long-document summarization (up to 1M tokens)
- High-volume classification and tagging
- Multimodal pipelines with image + text inputs
- Real-time chatbots needing sub-second latency
Not ideal for
- Complex multi-step reasoning tasks
- Tasks requiring maximum accuracy over speed
Real cost scenarios
| Scenario | Est. monthly cost | Breakdown |
|---|---|---|
| Document summarization — 10k docs/day × 5k tokens each | $750 | 50M tokens/day input at $0.50/M = $25/day |
| Chatbot — 100k conversations/month, 1k tokens avg | $150 | 100M input tokens at $0.50/M |
Estimates with typical token distributions. Use the Prompt Cost Calculator for your exact workload.
How to cut costs on Gemini 3 Flash
- Use Gemini 2.5 Flash-Lite at $0.10/M for the cheapest classification tasks — 80% cheaper than Gemini 3 Flash.
- Context caching is available — cache static document context and pay only $0.125/M for reads.
Calculate your Gemini 3 Flash costs
Other Google models
Compare Gemini 3 Flash
Frequently Asked Questions
What is Gemini 3 Flash context window?
1 million tokens — roughly 750,000 words or ~3,000 pages. This is 8× larger than GPT-5 mini (128K) and 5× larger than Claude Haiku 4.5 (200K). For long-document tasks, this eliminates chunking overhead entirely.
Is Gemini 3 Flash multimodal?
Yes — it processes text, images, audio, and video natively. For image-heavy pipelines, Gemini 3 Flash is competitively priced against GPT-4o Vision at $2.50/M input.
Pricing verified 2026-08-19 from Google official pricing. All prices in USD per 1M tokens. See methodology and price changelog.