GPT 6.1 Terra
Neural Network
GPT 6.1 Terra by OpenAI is a mid-range model for daily tasks, coding, and working with screenshots.
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, max
- Capabilities
- reasoning, function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
import OpenAI from 'openai';
const openai = new OpenAI({
apiKey: '<your bothub access token>',
baseURL: 'https://openai.bothub.chat/v1'
});
// Sync - Text generation
async function main() {
const chatCompletion = await openai.chat.completions.create({
messages: [{ role: 'user', content: 'Say this is a test' }],
model: 'gpt-6.1-terra',
});
}
// Async - Text generation
async function main() {
const stream = await openai.chat.completions.create({
messages: [{ role: 'user', content: 'Say this is a test' }],
model: 'gpt-6.1-terra',
stream: true
});
for await (const chunk of stream) {
const part: string | null = chunk.choices[0].delta?.content ?? null;
}
}
main();
How it works GPT 6.1 Terra?
GPT 6.1 Terra is a mid-range model in OpenAI's GPT-6.1 line, designed for daily agentic scenarios, programming, and text tasks, with strong screenshot understanding. In BotHub, it is available alongside over 250 other neural networks, allowing you to switch models for specific tasks without separate subscriptions. It works without VPN or foreign cards, with pay-as-you-go billing in rubles—you only pay for tokens used, and the balance does not expire. Companies get access to contracts, invoices, EDI, admin panels, and usage limits. Data is encrypted via AES-GCM and not stored. A unified OpenAI-compatible API allows integration into your own multi-model system. Practical use cases: analyzing interface screenshots, errors, or tables for clear answers; accelerating daily development (refactoring, debugging, code analysis); building multi-step agent chains; drafting and editing work texts; and generating standard reports or support response drafts.Frequently asked questions about GPT 6.1 Terra
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, accepts images and documents, can reason before answering, and call functions. A context window of up to 1,050,000 tokens allows loading large amounts of material, while an output of up to 128,000 tokens enables creating lengthy documents. It is suitable for file analysis, coding, and tasks requiring tools.
The model generates up to 128,000 tokens per response—a very large volume sufficient for long reports, articles, or code snippets. The input context reaches 1,050,000 tokens, allowing you to keep extensive materials in a single dialogue and refer back to them.
Yes, reasoning before answering is a capability of the model: it can break down a task into steps, verify intermediate conclusions, and then formulate the final result. This is useful for mathematics, logic, and multi-step instructions. This mode is valuable where the process of solving is as important as the final answer.
Yes, function calling and structured JSON output are supported. The model can access external tools and return answers in a specified format, which is convenient for integrations: connecting to databases, processing requests, and building pipelines. Through the BotHub OpenAI-compatible API, such logic can be implemented without complex setup.
Yes, in addition to text, the model accepts images and documents. You can upload a scan, PDF, or photo and ask it to analyze the content, compare files, or extract specific data. We do not have information regarding audio and video support, so please verify this before uploading such files.
We do not have specific quality metrics for the Russian language, so we cannot provide exact figures. The model works with documents and texts, and the context of up to 1,050,000 tokens allows loading entire materials. To determine if the results meet your needs, we recommend running a few of your real-world tasks through it.