Native Multimodal Reasoning
Seamlessly process text, images, audio, and video. Native integration prevents data loss found in traditional separate encoders.
gemini-3.5-flash / image-to-text
The Gemini 3.5 Flash API delivers a massive 1M token context window with native multimodal reasoning. Built for speed, this Gemini 3.5 model excels at video analysis, high-speed document QA, and low-latency agentic workflows at a fraction of the cost.
$ 0.9
$ 1.5
image
$ 5.4
$ 9
text
Key Gemini 3.5 Flash API Capabilities
The Gemini 3.5 Flash API is built for scale, speed, and intelligence. Leverage native multimodal processing and huge context for your next AI project.
High-Throughput Performance
Optimized for speed with 40-50% faster TTFT than Pro models. Ideal for real-time applications and high-frequency API calls.
Context Caching Savings
Drastically reduce costs for repetitive queries. Cache long prompts to save up to 90% on input costs for frequent datasets.
1M Token Context Window
Analyze massive datasets with near-perfect recall. Perfect for large document sets and complex codebases that require a broad view.
How to Get a gemini-3.5-flash API Key
Getting a gemini-3.5-flash API key takes four steps and a few minutes. Create a free GPTProto account, add credits, generate your key, and make your first call — at $0.9 / $5.4 it's a cheaper gemini-3.5-flash API key than going direct, and one key works across every model on the platform. Full gemini-3.5-flash Documentation is in the docs.
Sign up
Create your free GPT Proto account to begin. You can set up an organization for your team at any time.
Top up
Your balance can be used across all models on the platform, including gemini-3.5-flash, giving you the flexibility to experiment and scale as needed.
Generate your API key
In your dashboard, create an API key — you'll need it to authenticate when making requests to gemini-3.5-flash.
Make your first API call
Use your API key with our sample code to send a request to gemini-3.5-flash via GPT Proto and see instant AI-powered results.
Common Questions About Gemini 3.5 Flash API
Find technical details and usage tips for the Gemini 3.5 Flash API. We cover performance, cost-efficiency, and integration workflows for developers.
How does the Gemini 3.5 Flash API handle long context?
This Gemini 3.5 model uses a 1M token window with 99% recall. It processes massive datasets, like 10+ PDFs or entire codebases, without losing information. The Flash API architecture is specifically optimized to keep this process fast and cost-effective for developers using the API in production.
Can I use the Gemini 3.5 Flash API for video analysis?
Yes, it is natively multimodal. Unlike models that only see images, the Gemini 3.5 Flash API can reason across up to one hour of video frames natively. This makes it ideal for summarizing meetings or security footage without the need for manual frame sampling or complex pre-processing pipelines.
Is the Gemini 3.5 Flash API faster than Gemini 1.5 Pro?
Absolutely. This Gemini 3.5 variant is optimized for flash inference. It delivers a Time To First Token that is approximately 40-50% faster than the 1.5 Pro version. It is designed for high-throughput scenarios where low latency is critical for user experience, such as real-time customer support bots.
What are the costs for the Gemini 3.5 Flash API?
Pricing is very competitive: $0.075 per 1M input tokens for prompts under 128k. For larger contexts, a small surcharge applies. However, the Gemini 3.5 Flash API supports context caching, which can reduce repetitive query costs by up to 90%, making it one of the most economical high-context models available.
Does the Gemini 3.5 Flash API support JSON mode?
Yes, the API supports structured outputs via a specific response type. By setting the response format to JSON, developers can ensure the Gemini 3.5 output adheres to strict schemas. This is perfect for data extraction, entity recognition, and building reliable agentic workflows that require precise formatting.
How do I migrate my code to use Gemini 3.5 Flash?
Migration is seamless. Our API provides an OpenAI-compatible endpoint. You simply update your base URL and model name to point to Gemini 3.5 Flash. Most system prompts and function definitions from other providers are 1:1 compatible, allowing you to switch to this high-speed Gemini API in minutes.