API Reference

Excellent! Below is a detailed and well-structured API Reference section for the Uniport AI Docs, covering all the submenus with comprehensive content, developer-friendly guidance, and examples.

---

📑 API Reference

---

🔐 Authentication

Overview

All requests to the Uniport API must be authenticated using your API key. Authentication is done using the Authorization header with the Bearer schema.

How to Authenticate

Include your API key in every request:

Authorization: Bearer YOUR_API_KEY

Example with curl:

curl https://api.uniportai.site/v1/chat/completions \
  -H "Authorization: Bearer sk-abc123xyz" \
  -H "Content-Type: application/json" \
  -d '{...}'

⚠️ Keep your key secure. Do not expose it in frontend apps or public repositories.

---

🧠 Models & Capabilities

Uniport provides access to a wide range of AI models, including open-source and commercial models. Supported categories include:

Type	Examples	Use Cases
Text (Chat)	GPT-4, Claude 3, LLaMA-3	Assistants, chatbots, coding
Text (Completion)	Mistral, Falcon	Creative writing, summarization
Image	Stable Diffusion, DALLE-3	Art generation, design mockups
Embeddings	OpenAI, Cohere, BGE	Search, classification, clustering
Audio	Whisper, Bark, XTTS	Transcription, TTS, voice agents

Each model includes metadata such as:

• Cost per token

• Average latency

• Max context length

• Supported input/output format

You can browse model details in the Playground or via the /v1/models endpoint.

---

📈 Rate Limits

Default Limits

All accounts have default rate limits based on their plan:

Plan	RPM (Requests/Min)	TPS (Tokens/Sec)
Free	60 RPM	50 TPS
Pro	300 RPM	200 TPS
Enterprise	Custom	Custom

How Limits Work

• Rate limits are per API key.

• Exceeding limits returns: 429 Too Many Requests

Headers Returned:

X-RateLimit-Limit: 60
X-RateLimit-Remaining: 5
X-RateLimit-Reset: 1698671234

---

❌ Error Handling

Uniport returns standard HTTP status codes with informative error messages.

Common Errors

Code	Message	Reason
401	Unauthorized	Invalid or missing API key
403	Forbidden	API key disabled or lacks permissions
404	Not Found	Invalid endpoint or model ID
429	Too Many Requests	Rate limit exceeded
500	Internal Server Error	Server issue — try again later
400	Bad Request	Invalid input format or missing fields

Example Error Response:

{
  "object": "error",
  "message": "Invalid model specified",
  "type": "invalid_request_error"
}

---

📊 Usage & Billing

How Usage Is Tracked

Each request contributes to your billing:

• Text: Billed per 1,000 tokens

• Image: Billed per generation

• Audio: Billed per second of audio processed

• Embeddings: Billed per input unit (e.g., tokens or docs)

How to View Usage

• Log in to Dashboard

• Go to Billing → Usage Logs

• See token counts, requests per model, and estimated cost

Billing Intervals

• Billed monthly for Pro and Enterprise plans

• Free tier resets every 30 days

---

✅ Pro Tips

• Use streaming for chat models to reduce latency.

• Cache embeddings for repeated queries.

• Rotate models easily with minimal code changes thanks to Uniport's consistent API format.

• Compare real-time cost/performance in the Model Comparator.

---

Would you like me to proceed with content for the AI Endpoints section next (e.g., /v1/chat/completions, /v1/embeddings, etc.)?