gpt-oss-120B

Advanced open reasoning model with enterprise-grade capabilities.

About model

Enterprise-Ready Open Reasoning:
gpt-oss-120B delivers sophisticated chain-of-thought reasoning capabilities in a fully open model. Built with community feedback and released under Apache 2.0, this 120B parameter model provides transparency, customization, and deployment flexibility for organizations requiring complete data security & privacy control.

Quickstart guides

RAG

Building a RAG Workflow

Agents

Agent Workflows

Apps

Next.js Chat Quickstart

Performance benchmarks

Model	FrontierMath Tier 4	GPQA Diamond	HLE	SciCode	GDPval-AA	Terminal-Bench 2.1	Agent Arena	FrontierCode	DeepSWE
gpt-oss-120B		78.2%	18%	39%	15%	26%
Related open-source models
Competitor closed-source models
Claude Fable 5	87.8%	92.6%	53%	60%	62%	85%		53.5%	70%
Claude Opus 5	73.2%	93.2%	53%	56%	68%	89%	+16.4pp	53.4%	74%
GPT-5.6 Sol	82.9%	94.1%	47%	56%	61%	88%		47.5%	73%
Grok 4.5	24.4%	93.1%	40%	54%	51%	82%	+4.2pp	42.4%	54%
GPT-5.6 Luna	61.0%	91.1%	37%	53%	54%	81%		39.8%	67%

API usage

cURL
Python
Typescript

Endpoint:

openai/gpt-oss-120b

curl -X POST "https://api.together.xyz/v1/chat/completions" \
  -H "Authorization: Bearer $TOGETHER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-oss-120b",
    "messages": [
      {
        "role": "user",
        "content": "What are some fun things to do in New York?"
      }
    ]
}'

from together import Together

client = Together()

response = client.chat.completions.create(
  model="openai/gpt-oss-120b",
  messages=[
    {
      "role": "user",
      "content": "What are some fun things to do in New York?"
    }
  ]
)
print(response.choices[0].message.content)

import Together from 'together-ai';
const together = new Together();

const completion = await together.chat.completions.create({
  model: 'openai/gpt-oss-120b',
  messages: [
    {
      role: 'user',
      content: 'What are some fun things to do in New York?'
     }
  ],
});

console.log(completion.choices[0].message.content);

Model card
Architecture Overview:
• Mixture-of-Experts (MoE) architecture with SwiGLU activations
• Alternating attention layers between full context and sliding 128-token window
• Learned attention sink per-head for enhanced performance

Training Methodology:
• Comprehensive safety training and evaluation protocols
• Community feedback integration from global listening sessions
• Rigorous testing under Preparedness Framework
• Standard GPT-4o tokenizer with additional Harmony format tokens

Performance Characteristics:
• Native FP4 quantization for efficient inference
• 128K context window with RoPE positional encoding
• Chain-of-thought reasoning with adjustable effort levels
Applications & use cases
Enterprise Applications:
• Complex reasoning and analysis tasks
• Research and development support
• Technical documentation generation
• Strategic planning and decision support

Developer Use Cases:
• Code generation and review
• API development and integration
• System architecture design
• Technical troubleshooting and debugging

Industry Solutions:
• Healthcare: Clinical decision support and medical research
• Finance: Risk analysis and regulatory compliance
• Legal: Contract analysis and legal research
• Education: Curriculum development and tutoring systems

Deployment Scenarios:
• On-premises infrastructure for data sovereignty
• Private cloud deployments for security compliance
• Custom fine-tuning for domain-specific applications
• Multi-modal integration with existing systems