Gemini 3.8 Flash
Neural Network
Gemini 3.8 Flash is a Google model for code, agentic tasks, and multi-step reasoning, available via API.
Specifications
- Input
- text, images, documents
- Output
- text
- Context size
- 1 048 576 tokens
- Max answer length
- 65 536 tokens
- Reasoning effort
- low, medium, high
- Capabilities
- reasoning, function calling, Batch API
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
Image prompt
(per 1K tokens)
How it works Gemini 3.8 Flash?
Gemini 3.8 Flash is a Google Flash family model designed for fast responses without sacrificing reasoning quality. Compared to Gemini 3.7 Flash, it has improved in software development, agentic scenarios, and multi-step reasoning—where a chain of steps matters more than a single answer. In BotHub, the model is available from Russia without a VPN or foreign card; pay in rubles for tokens as you use them, and they don't expire. Nearby, in the same interface, are over 250 other models, making it easy to compare results, and a unified OpenAI-compatible API allows switching between them without rewriting your integration. Traffic is encrypted via AES-GCM, data is not stored, and teams have access to contracts, invoices, electronic document management, and an admin panel with limits. It is useful when you need to analyze someone else's code and suggest refactoring, build an agent with a sequence of actions, find the cause of a bug in logs, break down a complex task into steps, or process a stream of similar requests.Frequently asked questions about Gemini 3.8 Flash
You can use generated results for commercial purposes. You own all rights to the content you create. The only restriction: make sure your prompt does not include copyrighted third-party material. You are responsible for respecting the rights to any input data.
This is a Google text model with a 65,536-token context window. It accepts documents and images, reasons before answering, and calls functions. It is suitable for working with code, analyzing long materials, extracting data from files, and scenarios where a fast, structured response is needed.
The output limit is 65,536 tokens per response, the same as the context window. In practice, this is voluminous technical text, a large code module, or a detailed analysis of an entire document. If the response hits the ceiling, just ask it to continue in the next message.
Yes, the reasoning mode is confirmed: the model processes the task step-by-step and only then provides the result. This is noticeable in logic, mathematics, code debugging, and multi-step instructions. You get a ready-made answer, and the internal processing increases its accuracy on complex queries.
Yes, both. Function calling allows connecting external tools, search, and databases, while structured output provides the answer strictly according to a specified JSON schema. In BotHub, the model works via a unified OpenAI-compatible API, so changing models does not break your integration.
Images and documents are supported as input: the model natively reads files, parses tables, diagrams, and screenshots, and extracts data from them. There is no note about audio and video in our data, so you should check the required format in the BotHub interface or via the API.
Our data does not specifically record input and output languages, so we won't make any promises—it's faster to check with your own real task. The BotHub interface and payment with Russian cards work without a VPN, and if you are not satisfied with the result, there are over 250 other models in the same window.