Gemini 3.5 Flash Lite Batch
Neural Network
Gemini 3.5 Flash Lite is an efficient Google model for agentic tasks and sub-agent roles in multi-agent systems
Specifications
- Input
- text, images, documents
- Output
- text
- Context size
- 1 048 576 tokens
- Max answer length
- 65 536 tokens
- Reasoning effort
- minimal, low, medium, high
- Capabilities
- reasoning, function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
Image prompt
(per 1K tokens)
How it works Gemini 3.5 Flash Lite Batch?
Gemini 3.5 Flash Lite is a high-efficiency Google model with enhanced agentic capabilities. It is designed for sub-agents performing narrow, well-defined tasks within complex multi-agent processes where processing speed for high-volume, repetitive requests is critical. The batch version is designed for bulk job submission when volume is high and instant results are not required. Through BotHub, the model works from Russia without a VPN or foreign card; payment is in rubles based on actual token usage, and tokens do not expire. Over 250 models are available in one window: a unified OpenAI-compatible API allows switching tasks to another model without rewriting integrations, and chats are protected by AES-GCM encryption. For companies, we offer contracts, invoices, electronic document management (EDM), and an admin panel with limits. It is suitable if you are building a multi-agent system and assigning Flash Lite the role of executing individual steps, processing large arrays of requests or documents using a single template, building request routing between models, or testing hypotheses at scale before launch.Frequently asked questions about Gemini 3.5 Flash Lite Batch
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 1,048,576 token context: it analyzes large documents and images, writes and edits code, reasons before answering, and calls functions. It is suitable for processing large materials, data extraction, summarization, and integration into services via the unified BotHub API.
The model outputs up to 65,536 tokens per response — that's dozens of pages. Enough for a long article, detailed documentation, or a long code module. If the material is larger, break the task into parts: the million-token context allows you to keep the entire project in one dialogue.
Yes, the reasoning mode before answering is confirmed according to our data. The model builds a chain of steps before the final formulation, which is noticeable in logic, mathematics, and code debugging — where careful analysis of the condition is important, not just a quick response.
Yes. Function calling and structured JSON output are supported: the model will return data according to a specified schema and select the necessary tool itself. Convenient for agents, document parsing, and integrations. Through the OpenAI-compatible BotHub API, calls work just like in your current code.
According to our data, it accepts text, images, and documents as input, with documents handled natively — the model reads the entire file, not just extracted text. There is no mention of audio or video, so expect it to handle images and documents: diagrams, scans, tables, screenshots.
We do not have verified data on languages, so we won't make promises — the fastest way is to run your task directly in one window and compare it with another model. The service itself is Russian-speaking: access from Russia without a VPN or foreign card, payment based on actual token usage.