Skip to content

GPT-OSS 120B - Free Reasoning Model on Cloudflare, 128K Context

Model ID @cf/openai/gpt-oss-120b
Available on the Cloudflare Workers AI free plan, drawing from the 10,000 Neurons daily quota. The model is marked Zero data retention.

Model Information

PropertyValue
Context length128,000 tokens
Input price$0 (within free quota)
Output price$0 (within free quota)
ProviderOpenAI / Cloudflare
StatusAvailable on Workers Free
Input modalitiesText
Data retentionZero retention
Reasoning effortlow / medium (default) / high

Description

GPT-OSS 120B is OpenAI’s open-weight reasoning model, built for production-grade general-purpose workloads. Cloudflare’s hosted version supports function calling, reasoning effort, and batch processing, and is explicitly marked as zero data retention, making it suitable for scenarios sensitive to data retention.

Billing is calculated in Neurons: about 31,818 Neurons per million input tokens and 68,182 Neurons per million output tokens. The free quota is 10,000 Neurons per day, roughly equivalent to over a hundred thousand output tokens, depending on request length.

Usage Example

curl https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/v1/chat/completions \
  -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "@cf/openai/gpt-oss-120b",
    "messages": [{"role": "user", "content": "Explain the attention mechanism in Transformers"}]
  }'

Related Links

📚 Related Free Models

These models are also available for $0 on Cloudflare. Choose by context length and modality.

ModelContextDescription
GLM-4.7-Flash131KZhipu’s multilingual text model for chat, instruction
Qwen 3.8 27B262KAlibaba’s 27B multimodal model with text and image input,

📋 View all Cloudflare free models →

🔗 Site Navigation

Last updated on · Source: Cloudflare official documentation and pricing pages. Quotas and prices may change at any time.