GPT 6 Astra Pro
Neural Network
GPT 6 Astra Pro is an OpenAI model with a pro reasoning mode for complex analytical and engineering tasks
Specifications
- Input
- text, images, documents
- Output
- text
- Context size
- 1 050 000 tokens
- Max answer length
- 128 000 tokens
- Reasoning effort
- low, medium, high, extra high, max
- Capabilities
- reasoning, function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works GPT 6 Astra Pro?
GPT 6 Astra Pro is a version of OpenAI's GPT 6 Astra model that runs with a pro reasoning mode: it's the same base model, but with more resources for thinking and higher quality answers on complex tasks. If the standard version isn't enough for multi-step logic, this model analyzes the task more deeply before responding. In BotHub, it is available from Russia without a VPN or foreign card; payment is made with Russian cards in rubles and on a pay-as-you-go basis for consumed tokens that do not expire. Other models work in the same window, so you can compare answers and choose the one suitable for the task, and a unified OpenAI-compatible API allows you to switch to another model without rewriting the integration. Correspondence is encrypted with AES-GCM, and companies can connect via a contract with invoicing and EDI, an admin panel, and team limits. Use GPT 6 Astra Pro where reasoning accuracy is important: analyzing large requirements and documents, analytics with hypothesis testing, complex engineering and scientific tasks, reviewing and debugging complex code, and preparing technical materialsFrequently asked questions about GPT 6 Astra Pro
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.
The model works with text, images, and documents as input: it parses large files, analyzes data, writes and edits code, and prepares analytics and drafts. A context of 1,050,000 tokens allows you to upload voluminous material in its entirety — for example, reports or a codebase — and work with it in a single dialogue.
In a single response, the model outputs up to 128,000 tokens — this is enough for a long article, a chapter of documentation, or a large block of code without stitching parts together. The exact volume in words depends on the language and formatting, so rely specifically on the token limit.
Yes. The model can construct reasoning before the final answer, which helps with logic, mathematics, and code analysis tasks. This mode is convenient for parsing complex conditions, testing hypotheses, and step-by-step calculations. The answer arrives ready-made — no separate mode activation is required.
Yes, both scenarios are supported: function calling and structured JSON output. This means the model can be connected to external services and databases, and its answers can be parsed immediately by code without manual format editing. BotHub provides access to such models via a unified OpenAI-compatible API.
Images and documents — yes, they are accepted as input. There are no notes in our data regarding audio and video, so it is better to clarify this point with support. The model will analyze pictures and large files along with a text query — for example, for parsing a diagram or table.
We do not claim specific ratings for the Russian language: the model data does not specify the language of prompts and output, and we will not invent numbers. The easiest way to check the quality is on your own tasks — send a typical request and evaluate the result. You can ask questions in the language you are comfortable with.