Zephyr API
Zephyr API is a RESTful
interface to access a wide diversity of state-of-the-art language models. Designed
to deliver Pro-level
reasoning at Flash speeds with low cost, it's ideal for agentic workflows,
multi-turn chat, and coding assistance.
Key Features
- High speed and efficiency
- Pro-level reasoning
- Low operational cost
- Multimodal support (text, images)
- Temperature configuration
- SSE streaming
- Tool integration
Base URL
All API requests should be made to the following
base URL:
https://zephyr-api.matiasvergara.workers.dev/
Available Models
Zephyr API offers a wide range of
state-of-the-art models tailored to different performance and cost needs.
Standard Models
Available with regular API usage limits based on
your plan.
| Model ID |
Description |
Plan |
| gemini-3-flash |
Gemini 3.0 Flash (Flagship) |
Basic+
|
| gemini-2.5-flash-lite |
Gemini 2.5 Flash Lite |
Basic+
|
| gpt-5-nano |
GPT-5 Nano |
Basic+
|
| xai-grok-4-fast |
xAI Grok 4 Fast |
Pro+
|
| minimax-m2.1 |
MiniMax M2.1 |
Pro+
|
| glm-4.7 |
GLM 4.7 |
Mega+
|
| gpt-oss-120b |
GPT-OSS 120B |
Mega+
|
| deepseek-v3.2-thinking |
DeepSeek V3.2 with Thinking |
Mega
|
✨ Premium Models
Enterprise Exclusive — These
state-of-the-art models are only available with the Enterprise plan.
Daily Limit: Premium models
have a limit of 100
requests/day. Usage resets at midnight UTC.
| Model ID |
Description |
Plan |
| gpt-5.2 |
OpenAI GPT-5.2 |
Enterprise
|
| gemini-3-pro |
Gemini 3 Pro |
Enterprise
|
| opus-4.5 |
Claude Opus 4.5 |
Enterprise
|
| sonnet-4.5 |
Claude Sonnet 4.5 |
Enterprise
|
| haiku-4.5 |
Claude Haiku 4.5 |
Enterprise
|
| opus-4.5-reasoning |
Claude Opus 4.5 with Reasoning |
Enterprise
|
Feature Availability
Check which features are available in your plan.
| Feature |
Basic |
Pro |
Mega |
Enterprise |
Custom |
| Visual Input (Images) |
|
|
|
|
|
| Function Calling |
|
|
|
|
|
| URL Context & Web Search (Gemini models) |
|
|
|
|
|
Authentication
The Zephyr API uses API Keys to
authenticate requests and track usage against your plan limits.
Getting Your API Key
- Go to the Console section in the sidebar
- Your API key will be displayed in the API Key card
- Click the Copy button to copy your key
Using Your API Key
Include your API key in the
Authorization header of every
request:
Authorization: Bearer YOUR_API_KEY
Example Request
curl -X POST https://zephyr-api.matiasvergara.workers.dev/generate \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{"text": "Hello, world!"}'
Rate Limits & Usage
Your API key is linked to your subscription
plan. Usage limits include:
- Rate Limit: Requests per minute (varies by plan)
- Daily Tokens: Maximum tokens per day
- Daily Requests: Maximum requests per day
- Premium Models (Enterprise): 100 requests/day for premium
models
Usage resets at midnight UTC. Check the Console
for your current usage statistics.
Endpoints
1. Text Generation
Endpoint: POST /generate
Generates text using the selected language model.
Request
POST /generate
Content-Type: application/json
{
"text": "Your prompt here",
"temperature": 0.7,
"system": "You are a helpful assistant"
}
Parameters
| Parameter |
Type |
Required |
Default |
Description |
text |
string |
Yes |
- |
Your question or prompt |
temperature |
number |
No |
0.7 |
Controls creativity (0.0-1.0) |
system |
string |
No |
null |
System prompt |
image |
string |
No |
null |
Image URL for analysis |
url |
string |
No |
null |
Web page URL for context |
stream |
boolean |
No |
false |
Enable streaming |
history |
array |
No |
[] |
Conversation history |
2. Health Check
Endpoint: GET /health
Checks the API status.
GET /health
Response Format
Success Response
{
"success": true,
"model": "gemini-3-flash",
"message": "AI response here...",
"latency": 1.45,
"finish_reason": "STOP",
"usage": {
"prompt_tokens": 38,
"completion_tokens": 175,
"total_tokens": 213
},
"metadata": {
"request_id": "550e8400-e29b-41d4-a716-446655440000",
"timestamp": "2026-01-12T15:30:00.000Z",
"duration_ms": 1450,
"temperature_used": 0.7
}
}
Error Response
{
"error": "API Error",
"details": "Error message here",
"status": 500,
"request_id": "550e8400-e29b-41d4-a716-446655440000"
}
Status Codes
| Code |
Description |
| 200 |
Success |
| 400 |
Invalid parameters |
| 500 |
Server error |
Python Examples
Basic Usage
import requests
url = "https://zephyr-api.matiasvergara.workers.dev/generate"
response = requests.post(url, json={
"text": "What is artificial intelligence?",
"temperature": 0.7
})
print(response.json()["message"])
With System Prompt
response = requests.post(url, json={
"text": "Explain quantum physics",
"system": "You are a physics professor. Use simple language.",
"temperature": 0.5
})
print(response.json()["message"])
With Image
response = requests.post(url, json={
"text": "Describe this image",
"image": "https://example.com/photo.jpg"
})
print(response.json()["message"])
Streaming
response = requests.post(url, json={
"text": "Write a short story",
"stream": True
}, stream=True)
for line in response.iter_lines():
if line:
print(line.decode('utf-8'))
JavaScript Examples
Basic Usage
const response = await fetch("https://zephyr-api.matiasvergara.workers.dev/generate", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({
text: "What is artificial intelligence?",
temperature: 0.7
})
});
const data = await response.json();
console.log(data.message);
With System Prompt
const response = await fetch("https://zephyr-api.matiasvergara.workers.dev/generate", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({
text: "Explain quantum physics",
system: "You are a physics professor.",
temperature: 0.5
})
});
console.log((await response.json()).message);
Streaming
const response = await fetch(url, {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({ text: "Write a story", stream: true })
});
const reader = response.body.getReader();
const decoder = new TextDecoder();
while (true) {
const { done, value } = await reader.read();
if (done) break;
console.log(decoder.decode(value));
}
cURL Examples
Basic Request
curl -X POST https://zephyr-api.matiasvergara.workers.dev/generate \
-H "Content-Type: application/json" \
-d '{
"text": "What is artificial intelligence?",
"temperature": 0.7
}'
With System Prompt
curl -X POST https://zephyr-api.matiasvergara.workers.dev/generate \
-H "Content-Type: application/json" \
-d '{
"text": "Explain quantum physics",
"system": "You are a physics professor.",
"temperature": 0.5
}'
Health Check
curl https://zephyr-api.matiasvergara.workers.dev/health
Use Cases
Question & Answer
Simple factual queries and information
retrieval.
Creative Writing
Stories, poems, and creative content
generation.
Code Explanation
Analyze and explain code snippets.
Image Analysis
Describe and analyze images via URL.
Web Summarization
Summarize web pages and articles.
Multi-turn Chat
Conversational AI with history context.
Configuration
Temperature
Controls randomness in responses:
| Range |
Behavior |
| 0.0 - 0.3 |
Deterministic, factual, consistent |
| 0.4 - 0.7 |
Balanced (default: 0.7) |
| 0.8 - 1.0 |
Creative, varied, unpredictable |
System Prompts
Customize AI behavior with different personas:
# Professional tone
system = "You are a university professor"
# Casual tone
system = "You are talking to a 10-year-old"
# Technical tone
system = "You are a technical documentation writer"
Best Practices
1. Use Appropriate Temperature
Use 0.3 for facts, 0.7 for balanced, 0.9 for
creative tasks.
2. Provide Clear Instructions
Be specific: "Provide 5 facts about cats
focusing on behavior" vs "Tell me about cats"
3. Use System Prompts for Consistency
Define the AI's role and tone for consistent
responses.
4. Handle Errors Gracefully
Always wrap API calls in try/catch with proper
timeout handling.
5. Use Streaming for Long Responses
Enable streaming for stories, articles, and
long content.
API Status
Operational
https://zephyr-api.matiasvergara.workers.dev/