GPT 5.5 Pro Batch
Neural Network
GPT 5.5 Pro is an OpenAI model for deep reasoning, analyzing large documents, and coding, with a 922K input token context.
Specifications
- Input
- text, images, documents
- Output
- text
- Context size
- 1 050 000 tokens
- Max answer length
- 128 000 tokens
- Reasoning effort
- medium, high, extra high
- Capabilities
- reasoning, function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works GPT 5.5 Pro Batch?
GPT 5.5 Pro is an OpenAI model designed for tasks requiring accuracy and deep reasoning. It supports a context window of over a million tokens, accepts up to 922,000 input tokens, and returns up to 128,000 output tokens, allowing entire codebases, document sets, and multi-step calculations to fit into a single request. The batch version is for bulk processing when immediate responses aren't required. In BotHub, the model works from Russia without a VPN or foreign card; payment is in rubles per token used, and tokens do not expire. Access over 250 models in one window with easy switching and a unified OpenAI-compatible API, so no integration rewrite is needed. Requests are encrypted; corporate contracts, invoicing, and an admin panel with limits are available. Useful for analyzing large repositories to find bugs, performing reviews and refactoring, aggregating analytics across hundreds of documents, solving STEM problems with long derivation chains, or running mass text processing in batches.Frequently asked questions about GPT 5.5 Pro 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 text model with a reasoning mode: it analyzes long documents, writes and refactors code, and helps with analytics and STEM tasks. A 1,050,000 token context allows you to keep an entire repository or a collection of reports in a single request and work with them without splitting.
Up to 128,000 tokens per response—that's dozens of pages of text. Enough for a large technical document, detailed analysis, or an entire bulky code module without needing to ask it to 'continue'. If the task is even larger, break it into parts and assemble the result by section.
Yes, reasoning before answering is confirmed in our data. The model first builds a solution path and only then provides the result—this is noticeable in tasks involving math, logic, multi-step calculations, and code debugging, where it's important not just to guess the answer, but to derive it sequentially.
Yes, both function calling and structured JSON output are supported. The model can access your tools, databases, and services, and return answers in a specified schema—convenient for agents and pipelines. Connection is via the unified OpenAI-compatible BotHub API.
Images and documents—yes, they are listed as input formats: you can send a screenshot, diagram, scan, or PDF and get a text analysis. There is no mention of audio or video in our data, so please test such files with your task in the BotHub interface.
There is no specific note about languages in our data, so we won't promise specific accuracy—evaluate the quality with your own material. It's easy: in BotHub, access is available without a VPN or foreign card, payment is in rubles per token used, and tokens do not expire.