GPT OSS 120B Batch
Neural Network
GPT OSS 120B is an open OpenAI model with a Mixture-of-Experts architecture for reasoning and agentic tasks.
Max answer length
(in tokens)
Context size
(in tokens)
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works GPT OSS 120B Batch?
GPT OSS 120B is an open OpenAI language model with 117 billion parameters and a Mixture-of-Experts architecture: about 5.1 billion parameters are activated per pass, allowing the model to maintain high reasoning capabilities while remaining computationally efficient. It was designed for tasks requiring reasoning chains, agentic behavior, and stable production performance, with the batch variant optimized for processing large volumes of requests. Through BotHub, you can connect without a VPN or foreign card, pay with a Russian card in rubles for tokens used, and tokens do not expire. A unified OpenAI-compatible API allows you to integrate the model into your service and switch to any of the 250+ available models without rewriting the integration, while data is transmitted encrypted and not stored. It is useful for analyzing logs, explaining code errors, building agents with function calling and external tools, performing mass document classification, breaking down STEM problems step-by-step, or processing queues of similar requests at once.Frequently asked questions about GPT OSS 120B 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.
gpt-oss-120b:batch is a text model with a reasoning mode: it analyzes long documents, writes and refactors code, finds errors, solves math and logic problems, and prepares analytics and summaries. It has a 131,072 token context window, so large files and entire conversation threads fit into a request.
Up to 117,964 tokens in a single response — that's dozens of pages of text. The model easily outputs a large document, detailed analysis, or a bulky code module in its entirety without splitting it into parts. If the response cuts off, just ask it to continue from where it stopped.
Yes, reasoning before answering is a stated capability of the model: before formulating a result, it breaks down the condition step-by-step. This significantly helps with calculation, logic, and code debugging tasks. Note that reasoning also consumes output tokens.
Yes, the model supports function calling and structured JSON output — it is easy to integrate into agents, chatbots, and pipelines where the response needs to be parsed immediately. In BotHub, this works via a unified OpenAI-compatible API, so you can connect another model later without rewriting the integration.
Besides text, the model accepts documents — upload a file and ask for an analysis, summary, or answers based on the content. Working with images, audio, and video is not listed in its capabilities, so for such formats, switch to a specialized model in the same BotHub window.
We do not have data on the quality of performance for specific languages, so we won't make any promises — it's easier to test it on your own task. Payment is based on tokens used, and you can immediately send the same prompt to another model in the same window to compare responses.