Glimx Free Model Providers

This document lists free AI model providers that can be used with Glimx.

Configuration

Add these to your ~/.opencode/config.json or project's .opencode/opencode.jsonc:

{
  "provider": {
    "openrouter": {
      "api": "https://openrouter.ai/api/v1",
      "models": {
        "qwen-3-32b-free": {
          "id": "qwen/qwen3-32b:free",
          "cost": { "input": 0, "output": 0 },
          "limit": { "context": 32768, "output": 8192 }
        },
        "deepseek-chat-free": {
          "id": "deepseek/deepseek-chat-v3.1:free",
          "cost": { "input": 0, "output": 0 },
          "limit": { "context": 64000, "output": 8000 }
        },
        "mistral-small-free": {
          "id": "mistralai/mistral-small-3.2-24b-instruct:free",
          "cost": { "input": 0, "output": 0 },
          "limit": { "context": 32768, "output": 8192 }
        }
      }
    },
    "groq": {
      "api": "https://api.groq.com/openai/v1",
      "models": {
        "qwen-32b": {
          "id": "qwen/qwen3-32b",
          "cost": { "input": 0, "output": 0 },
          "limit": { "context": 32768, "output": 8192 }
        },
        "llama-3.3-70b": {
          "id": "llama-3.3-70b-versatile",
          "cost": { "input": 0, "output": 0 },
          "limit": { "context": 128000, "output": 32768 }
        }
      }
    },
    "cerebras": {
      "api": "https://api.cerebras.ai/v1",
      "models": {
        "qwen-235b": {
          "id": "qwen/qwen3-235b-a22b-instruct",
          "cost": { "input": 0, "output": 0 },
          "limit": { "context": 128000, "output": 8192 }
        },
        "llama-3.3-70b": {
          "id": "llama3.3-70b",
          "cost": { "input": 0, "output": 0 },
          "limit": { "context": 128000, "output": 8192 }
        }
      }
    }
  }
}

API Keys

Get your free API keys from:

• OpenRouter: https://openrouter.ai/keys (Free tier: 20 req/min)
• Groq: https://console.groq.com/keys (Free tier: 1000 req/day)
• Cerebras: https://cloud.cerebras.ai/console (Free tier available)

Set them in your environment:

export OPENROUTER_API_KEY="your-key-here"
export GROQ_API_KEY="your-key-here"
export CEREBRAS_API_KEY="your-key-here"

Recommended Free Models

For Coding (Best Performance)

1. Qwen 3 32B (Groq/OpenRouter) - Fast, excellent code generation
2. DeepSeek Chat V3.1 (OpenRouter) - Strong reasoning, code-focused
3. Llama 3.3 70B (Groq/Cerebras) - General purpose, very capable

For Thinking/Reasoning

1. Qwen 3 235B (Cerebras) - Massive context, deep thinking
2. DeepSeek R1 (OpenRouter) - Reasoning-focused architecture

For Speed

1. Groq - Ultra-fast inference (>500 tokens/sec)
2. Cerebras - Fast with large models

Usage

# Use specific model
glimx --model groq/qwen-32b

# Or set default in config
{
  "model": "groq/qwen-32b"
}

Rate Limits

Provider	Free Tier Limits
OpenRouter	20 req/min, 50 req/day
Groq	1000 req/day, 6000 tokens/min
Cerebras	30 req/min, 64K tokens/min

Notes

• All providers require API keys (free to obtain)
• Free tiers have rate limits but are generous for personal use
• Models update frequently - check provider docs for latest