Auto Router
Neural Network
OpenRouter's Auto Router automatically selects a model for your request — an auto router for text tasks.
Specifications
- Input
- text, images, documents
- Output
- text
- Context size
- 2 000 000 tokens
- Max answer length
- 660 000 tokens
- Reasoning effort
- low, medium, high
- Capabilities
- reasoning, function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works Auto Router?
Auto Router is not a separate neural network, but a router from OpenRouter: it analyzes your request and forwards it to the model best suited for the task. The selection is based on the collective experience of the OpenRouter community — which models users actually trust with their requests and budget. You don't need to manually compare options or remember which one handles what better. In BotHub, Auto Router works without a VPN or foreign card; payment is made with Russian cards in rubles based on actual usage, and unused tokens do not expire. Over 250 models are available in the same window, so you can switch to a specific one manually at any time. For development, there is a unified OpenAI-compatible API, so you won't have to rewrite the integration when changing models, and chats are encrypted with AES-GCM. Companies have access to contracts, invoices, electronic document management, and an admin panel with limits. Simple scenarios: daily work requests when there's no time to choose a model; drafts of emails, posts, and documents; code and technical text analysis; analytics and summary preparation; a test environment where you leave model selection to automation.Frequently asked questions about Auto Router
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.
auto is an automatic model selection mode: your request goes to a suitable neural network, so you don't need to switch manually. It accepts text, documents, and images as input, with a 200,000 token context. Suitable for working with long materials, code, analytics, and tasks requiring reasoning.
The output limit is 660,000 tokens per response, which is enough for very voluminous results: documentation, detailed analyses, and entire large code files. If generation hits the limit, just ask it to continue — the continuation will arrive in the next message in the same dialogue.
Yes, reasoning before answering is confirmed by our data: the model first breaks down the condition step-by-step and only then formulates the result. This is noticeable in math, logic problems, code debugging, and multi-step instructions where it's important not just to guess the answer, but to arrive at it consistently.
Yes, function calling and structured JSON output are supported. You describe the available tools, the model decides which one is needed, and returns arguments according to the specified schema. Through BotHub's unified OpenAI-compatible API, this integrates into agents and pipelines without rewriting the integration.
Images and documents — yes: according to our data, they are accepted as input, so you can analyze screenshots, diagrams, PDFs, and tables. There is no mention of audio or video in the data, so for transcribing recordings or analyzing videos, it is more convenient to switch to a specialized model in the same window.
We have no data on supported languages, so we won't provide an assessment. It is more reliable to test it with your own task: send a typical request, then switch to another model in the same window and compare the answers. Payment is based on actual token usage; no subscription is required.