API Reference

Basic health check endpoint. Returns service status and version information.

{
  "status": "healthy",
  "service": "tensorcortex-gateway",
  "version": "0.1.0",
  "environment": "production",
  "timestamp": "2025-01-02T12:00:00Z"
}

Generate chat completions using any supported model. OpenAI-compatible format for most providers (OpenAI, Groq, Mistral, Together, Fireworks, Perplexity, DeepInfra, OpenRouter). Anthropic and Google use their native formats.

{
  "model": "gpt-4o",
  "messages": [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Hello!"}
  ],
  "max_tokens": 1000,
  "temperature": 0.7,
  "stream": false
}

{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "model": "gpt-4o",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Hello! How can I help you today?"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 20,
    "completion_tokens": 10,
    "total_tokens": 30
  }
}

curl https://openai.tensor.cx/v1/chat/completions \
  -H "Authorization: Bearer your-openai-api-key" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o",
    "messages": [
      {"role": "user", "content": "Hello!"}
    ]
  }'

import OpenAI from 'openai'

// Just change baseURL - use your existing OpenAI SDK!
const client = new OpenAI({
  apiKey: 'your-openai-api-key',
  baseURL: 'https://openai.tensor.cx'
})

const response = await client.chat.completions.create({
  model: 'gpt-4o',
  messages: [{ role: 'user', content: 'Hello!' }]
})

console.log(response.choices[0].message.content)

List all available models for the selected provider. Response format matches provider's native format.

{
  "object": "list",
  "data": [
    {
      "id": "gpt-4o",
      "object": "model",
      "created": 1715367049,
      "owned_by": "system"
    },
    {
      "id": "gpt-4o-mini",
      "object": "model",
      "created": 1721172741,
      "owned_by": "system"
    }
  ]
}

Generate embeddings for text using various embedding models. Available on providers that support embeddings (OpenAI, Mistral, Together, etc.).

{
  "model": "text-embedding-3-large",
  "input": "The quick brown fox jumps over the lazy dog"
}

Model name (e.g., gpt-4o, claude-3-5-sonnet-20241022). Provider is determined by your base URL.

Array of message objects with role and content

Enable streaming responses via Server-Sent Events. Default: false

Cause: Malformed request body or invalid parameters (e.g., missing required fields, invalid JSON).

Fix: Validate your JSON syntax and ensure all required fields (model, messages) are present.

Cause: Invalid or missing API key. Key may be deleted, archived, or misspelled.

Fix: Verify your API key in the Cockpit dashboard. Check for extra whitespace or truncation.

Cause: Rate limit exceeded or cost limit reached (daily/monthly).

Fix: Implement exponential backoff. Check X-RateLimit-Reset header. Increase limits in Cockpit settings.

Cause: Gateway internal error. Rare occurrence, usually temporary.

Fix: Retry after a brief delay. Save the X-Request-Id and contact support if persistent.

Cause: Upstream provider API error (OpenAI, Anthropic, etc.). Invalid model name or provider issues.

Fix: Check provider status page. Verify model name is correct for the provider. Check your provider API key validity.

Cause: Gateway temporarily unavailable. Scheduled maintenance or high traffic.

Fix: Retry with exponential backoff. Check status page for announcements.