GPT 5.4 Batch
Neural Network
GPT 5.4 is an OpenAI model merging GPT and Codex: for code, reasoning, and working with a 922K token context.
Specifications
- Input
- text, images, documents
- Output
- text
- Context size
- 1 050 000 tokens
- Max answer length
- 128 000 tokens
- Reasoning effort
- none, low, medium, high, extra high
- Capabilities
- reasoning, function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works GPT 5.4 Batch?
GPT 5.4 is an OpenAI model merging GPT and Codex into one system, suitable for both dialogue and coding. The context exceeds a million tokens: 922K input and 128K output, allowing large repositories or documentation to fit in a request without cutting off long answers. The batch version is designed for bulk processing. Via BotHub, the model is available without VPN or foreign cards; payment is in rubles via Russian cards, paying only for tokens that do not expire. Over 250 models are available in one window, and a unified OpenAI-compatible API allows switching between them without rewriting integrations. Traffic is encrypted via AES-GCM, and companies have access to contracts, invoices, EDI, and an admin panel with limits. It assists developers with reviews and refactoring, analysts with large datasets, editors with long texts, and support teams with bulk request processing.Frequently asked questions about GPT 5.4 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.
It is a text model that accepts images and documents: code analysis and refactoring, debugging, analytics, STEM tasks, and working with long reports and specifications. A context of up to 1,050,000 tokens allows keeping entire projects and sets of files in one dialogue.
Up to 128,000 tokens in one response — enough for a large technical document, detailed analysis, or a complete code module without splitting. If the response hits the limit, just ask it to continue from where it stopped.
Yes, reasoning mode is supported: the model processes the task step-by-step before providing the result. This significantly helps in math, logic, architecture design, and debugging, where it is important to follow a chain of reasoning rather than just guessing the answer.
Yes, the model supports function calling and structured JSON output, making it easy to integrate into agents and services. BotHub provides a unified OpenAI-compatible API: you can switch to another model without rewriting your integration.
Images and documents — yes: the model natively analyzes interface screenshots, diagrams, charts, PDFs, and other files. Audio and video are not listed in our data, so it is better to choose a specialized model from the BotHub catalog for those formats.
We do not have specific data on languages, so it is best to test the model with your own phrasing and materials — it takes a couple of minutes. The BotHub interface is in Russian, payments are accepted via Russian cards, and no VPN or foreign cards are required.