SiliconFlow logo How to Get a Free SiliconFlow API Key (2026)

2 free models available — no credit card required. Get your SiliconFlow API key → Test free models →

SiliconFlow FreeLLM Score

45
🔹 Niche Provider — Consider for easy signup

Best suited for light testing or very specific narrow use cases.

🎁
Generosity Free limits
50/100
🌍
Accessibility Signup ease
65/100
📚
Breadth Model variety
25/100
Reliability Uptime
65/100
🔌
Compatibility Tool support
65/100
🧠
Quality Benchmarks
0/100

How we score →

What is SiliconFlow?

China-based free LLM API — DeepSeek, Qwen, BGE at 1,000 RPM.

SiliconFlow is a Chinese AI inference platform offering generous free tiers for DeepSeek, Qwen, BGE, and other models popular in the Chinese LLM ecosystem. The free tier supports up to 1,000 RPM and 50K TPM with no credit card required. Best latency for users in China and Asia-Pacific.

  • 1,000 RPM, 50K TPM free tier
  • DeepSeek, Qwen, BGE models
  • Lowest latency for China-based users
  • OpenAI-compatible endpoint

API Compatibility: OpenAI SDK-compatible (Chat Completions)

How to Get a SiliconFlow API Key

  1. 1
    Sign up at cloud.siliconflow.cn Phone verification. No credit card.
  2. 2
    Go to API Keys
  3. 3
    Create an API key
  4. 4
    Choose a model DeepSeek-V3, Qwen3, BGE available. 1,000 RPM free.
  5. 5
    Configure OpenAI client Base URL: https://api.siliconflow.cn/v1

All Free SiliconFlow Models — Context Windows & Rate Limits

Model Context Max Output Modality Rate Limit Released Status
Abbreviation 131K 8K text See provider page Online
deepseek-ai/DeepSeek-R1-Distill-Qwen-7B 131K 131K textreasoning 30 RPM, 60K TPM Online

SiliconFlow Free Tier Limits & Pricing

Credit Card Not required
Free Tier Permanently free
Context Range 131K – 131K
Total Models 2 free
Rate Limits 30 RPM, 60K TPM
API Compatibility OpenAI SDK-compatible (Chat Completions)

SiliconFlow API Setup Tutorial & Tools

SiliconFlow is fully compatible with popular AI coding assistants like Cursor, Claude Code, and more. To see step-by-step API configuration instructions for your favorite tool, please visit our Global Configuration Guide →

Use Cases

What SiliconFlow's free models are best for, based on aggregated model capabilities:

Chat 2 models Reasoning 1 model

Limitations & Caveats

  • China-hosted — high latency outside Asia-Pacific
  • Chinese phone number required for registration
  • Model availability may change based on local regulations

Frequently Asked Questions

Can overseas users register for SiliconFlow?

SiliconFlow currently requires a Chinese phone number for registration. International users may not be able to sign up without a Chinese number. Consider alternative providers like Groq or OpenRouter.

Is SiliconFlow faster than other providers for users in China?

Yes — SiliconFlow's servers are in mainland China, offering the lowest latency for Chinese users (typically under 100ms). This makes it ideal for China-based applications.

What models does SiliconFlow offer that others don't?

SiliconFlow specializes in Chinese ecosystem models: DeepSeek, Qwen, BGE embeddings, and Yi models. Some of these models are available elsewhere, but SiliconFlow often has the most up-to-date versions.

See our FAQ for common questions about free LLM APIs