{
  "opencollection": "1.0.0",
  "info": {
    "name": "Hugging Face Text Generation Inference API",
    "version": "1.0.0"
  },
  "request": {
    "auth": {
      "type": "bearer",
      "token": "{{bearerToken}}"
    }
  },
  "items": [
    {
      "info": {
        "name": "Text Generation",
        "type": "folder"
      },
      "items": [
        {
          "info": {
            "name": "Generate Text",
            "type": "http"
          },
          "http": {
            "method": "POST",
            "url": "https://api-inference.huggingface.co/generate",
            "body": {
              "type": "json",
              "data": "{}"
            }
          },
          "docs": "Generate text from a prompt using the loaded model. Returns the full generated text in a single response."
        },
        {
          "info": {
            "name": "Generate Text With Streaming",
            "type": "http"
          },
          "http": {
            "method": "POST",
            "url": "https://api-inference.huggingface.co/generate_stream",
            "body": {
              "type": "json",
              "data": "{}"
            }
          },
          "docs": "Generate text from a prompt using Server-Sent Events (SSE) streaming. Tokens are returned incrementally as they are generated."
        },
        {
          "info": {
            "name": "Create Text Completion (openai Compatible)",
            "type": "http"
          },
          "http": {
            "method": "POST",
            "url": "https://api-inference.huggingface.co/v1/completions",
            "body": {
              "type": "json",
              "data": "{}"
            }
          },
          "docs": "Create a text completion using the OpenAI-compatible completions API format."
        }
      ]
    },
    {
      "info": {
        "name": "Chat",
        "type": "folder"
      },
      "items": [
        {
          "info": {
            "name": "Create Chat Completion (openai Compatible)",
            "type": "http"
          },
          "http": {
            "method": "POST",
            "url": "https://api-inference.huggingface.co/v1/chat/completions",
            "body": {
              "type": "json",
              "data": "{}"
            }
          },
          "docs": "Create a chat completion using the OpenAI-compatible Messages API. Supports conversations, tool calling, structured output via JSON schema, and multi-modal inputs (text + images). Fully compatible with OpenAI client libraries."
        }
      ]
    },
    {
      "info": {
        "name": "Info",
        "type": "folder"
      },
      "items": [
        {
          "info": {
            "name": "List Available Models",
            "type": "http"
          },
          "http": {
            "method": "GET",
            "url": "https://api-inference.huggingface.co/v1/models"
          },
          "docs": "Get information about the model currently loaded on this TGI server."
        },
        {
          "info": {
            "name": "Get Server Information",
            "type": "http"
          },
          "http": {
            "method": "GET",
            "url": "https://api-inference.huggingface.co/info"
          },
          "docs": "Get metadata about the TGI server including the loaded model, version, SHA, Docker label, and model-specific parameters."
        },
        {
          "info": {
            "name": "Health Check",
            "type": "http"
          },
          "http": {
            "method": "GET",
            "url": "https://api-inference.huggingface.co/health"
          },
          "docs": "Check the health of the TGI server. Returns 200 if the model is loaded and ready to serve requests."
        },
        {
          "info": {
            "name": "Get Prometheus Metrics",
            "type": "http"
          },
          "http": {
            "method": "GET",
            "url": "https://api-inference.huggingface.co/metrics"
          },
          "docs": "Get server metrics in Prometheus format including request counts, latencies, queue sizes, and token throughput."
        },
        {
          "info": {
            "name": "Tokenize Text",
            "type": "http"
          },
          "http": {
            "method": "POST",
            "url": "https://api-inference.huggingface.co/tokenize",
            "body": {
              "type": "json",
              "data": "{}"
            }
          },
          "docs": "Tokenize a string of text using the model's tokenizer. Returns the token IDs and their string representations."
        }
      ]
    }
  ],
  "bundled": true
}