Providers Setup

🔌 Providers & API Setup Guide

This guide provides step-by-step instructions for obtaining API keys, configuring endpoints, and setting up all supported cloud and local LLM providers in Traliran AI Hub.

---

☁️ Cloud AI Providers

1. OpenAI

• Get API Key: Sign up or log in at https://platform.openai.com/api-keys and create a new secret key.
• Base Endpoint: https://api.openai.com/v1
• Supported Models: all models
• Notes: Requires an active paid platform account or pre-funded credit balance.

2. Anthropic Claude

• Get API Key: Generate a key in the Anthropic Console at https://console.anthropic.com/settings/keys.
• Base Endpoint: https://api.anthropic.com/v1
• Supported Models: all models
• Notes: Browser requests directly to Anthropic may require handling the anthropic-dangerous-direct-browser-headers flag depending on API policies.

3. Google Gemini

• Get API Key: Obtain a free or paid API key via Google AI Studio at https://aistudio.google.com/app/apikey.
• Base Endpoint: https://generativelanguage.googleapis.com
• Supported Models: all models
• Notes: Gemini provides generous free tier limits for developers.

4. Groq

• Get API Key: Create a key at https://console.groq.com/keys.
• Base Endpoint: https://api.groq.com/openai/v1
• Supported Models: all models
• Notes: High-speed inference engine using LPU chips. Fully OpenAI-compatible format.

5. DeepSeek

• Get API Key: Access the DeepSeek Platform at https://platform.deepseek.com/api_keys.
• Base Endpoint: https://api.deepseek.com/v1
• Supported Models: all models
• Notes: Extremely cost-effective API with full OpenAI schema compatibility.

6. OpenRouter

• Get API Key: Register at https://openrouter.ai/keys.
• Base Endpoint: https://openrouter.ai/api/v1
• Supported Models: all models
• Notes: Acts as an aggregator for hundreds of open and proprietary models with a single key.

7. GLM (Zhipu AI)

• Get API Key: Obtain an API key from Zhipu AI Open Platform at https://open.bigmodel.cn/usercenter/apikeys.
• Base Endpoint: https://open.bigmodel.cn/api/paas/v4
• Supported Models: all models
• Notes: Provides fast inference and high multilingual capabilities.

8. Qwen (Alibaba DashScope)

• Get API Key: Get a DashScope API key at https://dashscope.console.aliyun.com/apiKey.
• Base Endpoint: https://dashscope.aliyuncs.com/compatible-mode/v1
• Supported Models: all models
• Notes: Uses Alibaba Cloud's OpenAI-compatible API endpoint format.

---

💻 Local & Self-Hosted Providers

9. Ollama

• Installation: Download and install Ollama from https://ollama.com.

• Default Endpoint: http://localhost:11434

• Configuration (CRITICAL - CORS Setup): By default, Ollama blocks web applications running on different domains/ports. You must set the OLLAMA_ORIGINS environment variable before launching Ollama.

  Linux / macOS: export OLLAMA_ORIGINS="*" ollama serve

  Windows (PowerShell): $env:OLLAMA_ORIGINS="*" ollama serve

• Supported Models: Any model pulled locally via command line (e.g., ollama run llama3.2, ollama run gemma2).

10. llama.cpp (llama-server)

• Installation: Build or download pre-compiled binaries from https://github.com/ggerganov/llama.cpp.

• Default Endpoint: http://localhost:8080

• Configuration (CRITICAL - CORS Setup): Launch the llama-server executable with the --host and --cors flags enabled:

  ./llama-server -m your_model.gguf --host 0.0.0.0 --port 8080 --cors "*"

• Notes: Native C/C++ execution for GGUF model files with minimal memory overhead.