🤖Models & Info

Query the /models endpoint with a GET request to get information about available online models including their specifications and capabilities.

curl https://api.modelsync.ai/v1/models

Response format:

{
    "models": [{
        "id": "Llama-3.3-70B-Instruct",
        "name": "Llama-3.3-70B-Instruct",
        "type": "text",
        "creator": "meta-llama",
        "parameter_size": "70B",
        "size": "large",
        "credit_cost": 2,
        "hf_link": "https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct",
        "quantization": "awq",
        "context_length": 8192,
        "max_completion_tokens": 400
    }, {
        "id": "Llama-3.1-Nemotron-70B-Instruct-HF",
        "name": "Llama-3.1-Nemotron-70B-Instruct-HF",
        "type": "text",
        "creator": "nvidia",
        "parameter_size": "70B",
        "size": "large",
        "credit_cost": 2,
        "hf_link": "https://huggingface.co/nvidia/Llama-3.1-Nemotron-70B-Instruct-HF",
        "quantization": "awq",
        "context_length": 8192,
        "max_completion_tokens": 400
    }, ...
    ]
}

⚡Using Models

When making requests, specify the model using its ID in the model parameter:

{
  "model": "Llama-3.3-70B-Instruct",
  "messages": [
    {"role": "user", "content": "Hello!"}
  ]
}

Tips for Best Results

Test different models to find the best fit for your specific use case - each model has its own strengths
Consider context_length (prompt tokens input + completion tokens output) requirements when choosing a model
Check model details on Hugging Face using the hf_link to understand model-specific parameters and capabilities
Monitor response times and credit costs to find the optimal balance for your usage