GPT 6.1 Terra Pro
Neural Network
GPT-6.1 Terra Pro by OpenAI is a Terra version with a pro mode for higher quality work on complex 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 6.1 Terra Pro?
GPT-6.1 Terra Pro by OpenAI is the same GPT-6.1 Terra model, but with reasoning pro mode enabled: it processes requests more deeply, providing higher quality results where multiple steps and careful conclusions are needed. Essentially, it's a way to get a thoughtful response from a familiar model without rebuilding your integration: in BotHub, Pro and standard Terra are side-by-side in one window, allowing you to compare answers in a few clicks — BotHub offers over 250 models. No subscription is required — pay-as-you-go for tokens that don't expire. You can work from Russia without a VPN or foreign card, using a Russian card. For multi-model systems, a unified OpenAI-compatible API is available; for teams, we offer contracts, invoices, EDI, limits, and an admin panel. Note that pro mode consumes more tokens, so enable it for specific tasks. Terra Pro is suitable where the cost of error outweighs speed: analytics, decision justification, hypothesis and calculation testing, design, preparing detailed reports, and final proofreading of responses generated in standard mode.Frequently asked questions about GPT 6.1 Terra 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 understands text, images, and documents, can reason before answering, and call functions. A context window of up to 1,050,000 tokens allows loading large materials: code, reports, tables. It is suitable for analytics, development, documentation analysis, and data extraction, and you can switch to another model without rewriting your integration.
The model generates up to 128,000 tokens per response — enough for long materials: chapters, voluminous reports, large code fragments. If you need more, break the task into parts and pass the previous result in the next request: the 1,050,000 token context allows this.
Yes, reasoning before answering is a confirmed capability of the model: it goes through steps before providing a result. This helps where logic and condition checking are important — in calculations, code analysis, and debugging. You get a finished answer, not a draft.
Yes, function calling and structured JSON output are confirmed capabilities. It is easy to integrate into scenarios requiring a strict format or external service calls, including via BotHub's unified OpenAI-compatible API: you won't need to rewrite the integration when switching models.
Images and documents — yes, the model accepts them. We have no notes on audio and video in our data, so we cannot promise: if this is critical for your scenario, please check with support. It is convenient to pass an image as a reference and a document as a source of facts for analysis.
There is no separate evaluation for the Russian language in our data, so we cannot provide specific figures. It is more reliable to test it with your own tasks: give the model a typical request with your terminology and see if the result is satisfactory. Payment is pay-as-you-go for tokens, without a subscription.