AI Developer Environment & Tool Config Hub
Generate copy-paste configuration snippets for VS Code Continue, Cline, Cursor, Claude Code, OpenCode, Ollama, and vLLM. Connect verified free LLM backends with no credit card required.
Claude Code — Free Model Configuration
Configure Claude Code with free backends via ANTHROPIC_BASE_URL and ANTHROPIC_AUTH_TOKEN environment variables. Works with OpenRouter, NVIDIA NIM, and SambaNova.
Quick Start
Claude Code — OpenRouter Setup
# Get API Key from OpenRouter Open: https://openrouter.ai/keys # Anthropic's official CLI coding agent. Uses ANTHROPIC_BASE_URL and ANTHROPIC_AUTH_TOKEN environment variables to connect to any Anthropic-compatible backend.
Install Claude Code — npm install -g @anthropic-ai/claude-code
Set environment variables in your shell profile — ~/.bashrc or ~/.zshrc
Restart Claude Code and verify connection — claude --version
Environment Variables
ANTHROPIC_BASE_URLREQUIREDexport ANTHROPIC_BASE_URL=https://openrouter.ai/api/v1ANTHROPIC_AUTH_TOKENREQUIREDexport ANTHROPIC_AUTH_TOKEN=sk-or-v1-xxxxxxxxxANTHROPIC_API_KEYexport ANTHROPIC_API_KEY=""Cursor — Custom API Endpoint Configuration
Set up Cursor with custom OpenAI-compatible API endpoints. Add custom models via Settings → Models → Add Custom Model using Groq, Google AI Studio, or SambaNova free tiers.
Quick Start
Cursor — Groq Setup
# Get API Key from Groq Open: https://console.groq.com/keys # AI-powered code editor with built-in chat and diff editing. Supports custom OpenAI-compatible endpoints through Settings → Models → Add Model.
Open Cursor Settings — Ctrl+, or Cmd+,
Navigate to Models → Add Custom Model — Models section in Settings
Enter Base URL, API Key, and Model ID — Override OpenAI Base URL field
Save and start coding with free models — Chat panel
Environment Variables
OPENAI_BASE_URLREQUIREDhttps://api.groq.com/openai/v1OPENAI_API_KEYREQUIREDgsk_xxxxxxxxx or sk-xxxxxxxxxMODEL_IDREQUIREDmeta-llama/llama-3.3-70bOpenCode — JSON Configuration Guide
Configure OpenCode AI coding agent via ~/.config/opencode/opencode.json. Add any OpenAI-compatible provider including Groq, SambaNova, and OpenRouter with no credit card.
Quick Start
OpenCode — Groq Setup
# Get API Key from Groq Open: https://console.groq.com/keys # Open-source AI coding agent with a JSON configuration file. Configure OpenAI-compatible providers through ~/.config/opencode/opencode.json.
Install OpenCode — npm install -g opencode
Create or edit ~/.config/opencode/opencode.json — JSON config file
Add provider configuration with baseURL and apiKey — providers array
Restart OpenCode and select the model — Model selector
Environment Variables
OPENAI_BASE_URLREQUIREDhttps://api.sambanova.ai/v1OPENAI_API_KEYREQUIREDsb-xxxxxxxxxVS Code Continue — Free LLM Setup Guide
Configure VS Code Continue extension with free LLM backends. Edit ~/.continue/config.json to connect Groq, SambaNova, OpenRouter, or NVIDIA NIM with no credit card required.
Quick Start
VS Code — Continue — Groq Setup
# Get API Key from Groq Open: https://console.groq.com/keys # VS Code extension for AI-assisted coding. Configure via ~/.continue/config.json with any OpenAI-compatible or Anthropic-compatible backend.
Install Continue extension — VS Code Extensions → 'Continue'
Open Continue settings — Ctrl+, → Extensions → Continue
Edit ~/.continue/config.json — ~/.continue/config.json
Add provider config and restart VS Code — provider + model
Environment Variables
OPENAI_BASE_URLREQUIREDhttps://api.groq.com/openai/v1OPENAI_API_KEYREQUIREDgsk_xxxxxxxxxMODELmeta-llama/llama-3.3-70bVS Code Cline — OpenAI Compatible Backend Setup
Set up VS Code Cline extension with OpenAI-compatible backends. Configure OPENAI_BASE_URL and OPENAI_API_KEY to use free models from Groq, SambaNova, and OpenRouter.
Quick Start
VS Code — Cline — Groq Setup
# Get API Key from Groq Open: https://console.groq.com/keys # VS Code extension for autonomous AI coding. Uses OPENAI_BASE_URL and OPENAI_API_KEY environment variables for any OpenAI-compatible backend.
Install Cline extension — VS Code Extensions → 'Cline'
Open Cline settings — Ctrl+, → Extensions → Cline
Set provider to OpenAI Compatible — Provider dropdown
Enter Base URL and API Key — Settings panel
Environment Variables
OPENAI_BASE_URLREQUIREDhttps://api.groq.com/openai/v1OPENAI_API_KEYREQUIREDsk-xxxxxxxxxMODEL_IDmeta-llama/llama-3.3-70bVS Code Roo Code — Free AI Backend Setup
Configure VS Code Roo Code with custom API providers. Add OPENAI_BASE_URL and OPENAI_API_KEY to connect to free LLM backends from Groq and SambaNova.
Quick Start
VS Code — Roo Code — Groq Setup
# Get API Key from Groq Open: https://console.groq.com/keys # VS Code extension for AI-powered coding with multiple model providers. Configure via ~/.roo/config.json or VS Code settings.
Install Roo Code extension — VS Code Extensions → 'Roo Code'
Open Roo Code settings — Ctrl+, → Extensions → Roo Code
Add custom API provider — Provider settings
Enter Base URL and API Key — Settings panel
Environment Variables
OPENAI_BASE_URLREQUIREDhttps://api.groq.com/openai/v1OPENAI_API_KEYREQUIREDsk-xxxxxxxxxOllama — Local LLM Runner Free Setup
Run LLMs locally with Ollama. No API keys or credit cards required. Pull models like Llama 3.3, Qwen 2.5, and DeepSeek R1 for free local inference.
Quick Start
Ollama — Custom Setup
# Get API Key from provider # No API key required for local backends # Local LLM runner. Ollama runs models locally with no API keys required. For remote backends, configure OLLAMA_BASE_URL.
Install Ollama — ollama.com/install
Pull a model — ollama pull llama3.3
Run locally — ollama run llama3.3
Environment Variables
OLLAMA_BASE_URLhttp://localhost:11434OLLAMA_HOST0.0.0.0:11434vLLM — High-Throughput Local Serving Setup
Deploy vLLM server locally for high-throughput LLM inference. Start with vllm serve command and connect any OpenAI-compatible tool to http://localhost:8000/v1.
Quick Start
vLLM — Custom Setup
# Get API Key from provider # No API key required for local backends # High-throughput LLM serving engine. Configure vLLM server with any OpenAI-compatible model and connect any tool via the local server URL.
Start vLLM server — vllm serve model-name --host 0.0.0.0 --port 8000
Use server URL in any tool — http://localhost:8000/v1
Environment Variables
VLLM_BASE_URLREQUIREDhttp://localhost:8000/v1OPENAI_API_KEYdummy-keyCodex / Aider — Free OpenAI Compatible Setup
Configure OpenAI Codex, Aider, or Continue with custom OPENAI_BASE_URL and OPENAI_API_KEY. Connect to NVIDIA NIM, Groq, or OpenRouter free tiers.
Quick Start
Codex / Aider / Continue — NVIDIA NIM Setup
# Get API Key from NVIDIA NIM Open: https://build.nvidia.com/ # OpenAI Codex, Aider, and Continue all support OpenAI-compatible API endpoints via environment variables or configuration files. Works with any provider exposing an OpenAI-compatible API.
Set OPENAI_BASE_URL and OPENAI_API_KEY — Shell environment
Configure model in aide.conf.json or config.toml — Config file
Run Codex/Aider/Continue normally — CLI command
Environment Variables
OPENAI_API_KEYREQUIREDsk-xxxxxxxxxOPENAI_BASE_URLREQUIREDhttps://integrate.api.nvidia.com/v1OPENAI_MODELdeepseek-ai/deepseek-r1Gemini CLI — Google AI Studio Free Setup
Configure Google's Gemini CLI tool with GEMINI_API_KEY from Google AI Studio. Access Gemini 2.0 Flash and 2.5 Flash with generous free quotas.
Quick Start
Gemini CLI — Google AI Studio Setup
# Get API Key from Google AI Studio Open: https://aistudio.google.com/apikey # Google's official Gemini CLI tool. Uses GEMINI_API_KEY environment variable to authenticate with Google AI Studio.
Install Gemini CLI — npm install -g @google/gemini-cli
Set GEMINI_API_KEY in your shell profile — ~/.bashrc or ~/.zshrc
Restart Gemini CLI and verify connection — gemini --version
Environment Variables
GEMINI_API_KEYREQUIREDAIzaSyxxxxxxxxxPython OpenAI SDK — Free API Configuration
Drop-in Python OpenAI SDK setup with custom baseURL. Connect to Groq, SambaNova, Google AI Studio, or OpenRouter free tiers with zero code changes.
Quick Start
Python OpenAI SDK — Groq Setup
# Get API Key from Groq Open: https://console.groq.com/keys # Drop-in Python SDK configuration with custom baseURL. Works with any OpenAI-compatible provider without code changes.
Environment Variables
OPENAI_BASE_URLREQUIREDhttps://api.groq.com/openai/v1OPENAI_API_KEYREQUIREDgsk_xxxxxxxxxNode.js OpenAI SDK — Free Backend Configuration
Drop-in Node.js OpenAI SDK setup with custom baseURL. Connect to Groq, SambaNova, or OpenRouter free tiers with zero code changes.
Quick Start
Node.js OpenAI SDK — Groq Setup
# Get API Key from Groq Open: https://console.groq.com/keys # Drop-in Node.js OpenAI SDK configuration with custom baseURL. Works with any OpenAI-compatible provider without code changes.
Environment Variables
OPENAI_BASE_URLREQUIREDhttps://api.groq.com/openai/v1OPENAI_API_KEYREQUIREDsk-xxxxxxxxxVerified Free Backend Providers
All providers below support OpenAI-compatible API endpoints. No credit card required for any of them.
Google AI Studio
No CardGoogle's official AI studio. Gemini 2.0 Flash and 2.5 Flash with generous free quotas.
gemini-2.0-flashgemini-2.5-flashgemma-2-9bGroq
No CardUltra-low latency LPU inference. Fastest free tier at 330+ tok/s on Llama 3.3 70B.
meta-llama/llama-3.3-70bmeta-llama/llama-3.1-8bqwen/qwen2.5-coder-32bSambaNova Systems
No CardReconfigurable Dataflow Unit (RDU) inference. Sub-second latency with daily free quota.
meta-llama/llama-3.3-70bdeepseek-ai/deepseek-r1qwen/qwen2.5-72bNVIDIA NIM
No CardEnterprise-grade GPU-accelerated inference microservices with free promotional credits.
meta-llama/llama-3.3-70bnvidia/nemotron-nanomistralai/mistral-largeOpenRouter
No CardMulti-provider routing layer aggregating free model endpoints. :free models available.
meta-llama/llama-3.3-70b:freeqwen/qwen2.5-coder-32b:freedeepseek-ai/deepseek-r1:freeMistral AI
No CardEuropean open-source LLM platform with MoE architecture. Free tier requires phone verification.
mistralai/mistral-nemo-12bmistralai/codestral-22bmistralai/mistral-large-latestGitHub Models
No CardFree AI models via GitHub Models with GitHub account. GPT-4o mini, Llama 3, Phi-4 available.
gpt-4o-minimeta-llama/llama-3-8bmicrosoft/phi-4Cerebras
No CardWafer-scale AI inference on custom CS-3 systems. Fastest free tier at 2,100+ tok/s on Llama 3.3 70B.
meta-llama/llama-3.3-70bmeta-llama/llama-3.1-8b