> ## Knowledge Base Index
> Fetch the complete knowledge base index at: https://help.janitorai.com/sitemap.xml
> Use this file to discover available pages before exploring further.
> Pure-Markdown content can be obtained by appending a '.md' suffix to the content URLs listed in the sitemap (without the trailing slash).

# Troubleshooting Deepseek via Openrouter

If you're encountering errors when using DeepSeek models through OpenRouter, here’s a breakdown of common error codes, their causes, and how to fix them.

### **ERROR 400: Bad Request**
**Cause**: Incorrect proxy settings.
**Fix**: Double-check the model name, proxy URL, and API key for any extra spaces or typos.
**Correct Model Names**:
* **R1 Free**: `deepseek/deepseek-r1:free`
* **R1 Paid**: `deepseek/deepseek-r1`
* **V3 Free**: `deepseek/deepseek-chat-v3-0324:free`
* **V3 Paid**: `deepseek/deepseek-chat-v3-0324`
* **Legacy V3 Free**: `deepseek/deepseek-chat:free`
**Proxy URL**:
`https://openrouter.ai/api/v1/chat/completions`

### **ERROR 401: Invalid Credentials**
**Cause**: Problem with your API key.
**Fix**:
* Generate a new API key.
* Ensure you're using it on the same device it was created on.
* Make sure **“Model Training”** is set to **"On"** in OpenRouter settings for free models.

### **ERROR 402: Insufficient Credits**
**Cause**: You’ve run out of paid credits.
**Fix**:
* Add more credits to your OpenRouter account.
* Or switch to a **Free** version of the DeepSeek model.

### **ERROR 429: Rate Limited**
**Cause**: Too many requests.
**Fix**:
* **Daily limits**:
    * Free users with **< $10 credits**: 50 messages/day.
    * Free users with **$10+ credits**: 1000 messages/day.
* **Message cap**: Max **10 messages per minute** (resets every 3 hours).
* **Wait** for your quota to reset at **12AM GMT**.
* If error log includes `ErrorUpstreamFault:True`, DeepSeek servers may be down.

### **ERROR 500: Server Error**
**Cause**: DeepSeek servers are having issues.
**Fix**: Try again later — typically due to high traffic or outages.

### **ERROR 502: Invalid Response / Model Down**
**Cause**: The selected model is down or returned an invalid response.
**Fix**:
* Wait and try again later.
* Check model uptime at:
👉 [DeepSeek R1 Free Uptime](https://openrouter.ai/deepseek/deepseek-r1:free/uptime)