GPT-OSS 120B

OpenAI GPT-OSS 120B is an open-weight Mixture-of-Experts model (117B total, 5.1B active per forward pass) with configurable reasoning depth, native tool use, browsing, function calling, and structured outputs. Runs efficiently on a single H100 GPU thanks to native MXFP4 quantization.

ReasoningToolsopen weights

Context 131K·Max out 131K

Quickstart

import { streamText } from 'ai';
import { createOpenAICompatible } from '@ai-sdk/openai-compatible';

const blackbox = createOpenAICompatible({
  name: 'blackbox',
  apiKey: process.env.BLACKBOX_API_KEY!,
  baseURL: 'https://api.blackbox.ai/v1',
});

const result = streamText({
  model: blackbox('blackboxai/openai/gpt-oss-120b'),
  prompt: 'Why is the sky blue?',
});

for await (const chunk of result.textStream) {
  process.stdout.write(chunk);
}

Specs & pricing

Input $/M: 0.35
Output $/M: 0.75
Context window: 131K
Max output: 131K
Supported parameters: tool-usereasoningjson-modestreaming

Related models

Model	Type	Context	Input $/M
Blackbox Pro	Text	400K	1.75
GPT-5.5 Pro	Text	1M	30
GPT-5.5	Text	1M	5
GPT-5.4 Pro	Text	1.1M	30
GPT-5.4	Text	1.1M	2.50