Bestseller #2
Bestseller #3
Bestseller #4
Bestseller #6
Bestseller #7
AI Export
All systems operational
API Reference · REST · 2025
AI Export
API Layer
A unified inference interface to the world’s most capable language models. Build production-grade applications with streaming, batching, and structured outputs.
BASE URL
https://api.aiexport.io/v2
Uptime
99.98%
30-day average
Median latency
340ms
p50, streaming first token
Models available
14
Across 4 families
API version
2.4
Released Apr 2025
01
Authentication
API Key
Pass your API key in the X-API-Key header with every request. Keys are scoped per workspace and can be rotated without downtime.
# Every request needs this header
curl https://api.aiexport.io/v2/messages \
-H “X-API-Key: axp_live_••••••••••••” \
-H “Content-Type: application/json”
OAuth 2.0 (Enterprise)
Enterprise accounts may use OAuth 2.0 client credentials flow for server-to-server auth. Tokens expire after 3,600 seconds and must be refreshed.
Note: Never expose API keys in client-side code or public repos. Use environment variables or a secrets manager.
02
Models
Apex Opus 3
apex-opus-3-20250401
Context
200K
Input
$15 / 1M
Output
$75 / 1M
Vision
Yes
Apex Sonnet 4
apex-sonnet-4-20250501
Context
200K
Input
$3 / 1M
Output
$15 / 1M
Vision
Yes
Apex Haiku 4
apex-haiku-4-20250401
Context
100K
Input
$0.25 / 1M
Output
$1.25 / 1M
Vision
Yes
Fine-tuned (Custom)
ft-{workspace}-{model-id}
Context
Inherited
Input
+$1 / 1M
Output
+$5 / 1M
Vision
Conditional
03
Endpoints
POST
/messages
Create a message
Core
const response = await fetch(“https://api.aiexport.io/v2/messages”, {
method: “POST”,
headers: {
“X-API-Key”: process.env.API_KEY,
“Content-Type”: “application/json”
},
body: JSON.stringify({
model: “apex-sonnet-4-20250501”,
max_tokens: 1024,
messages: [{ role: “user”, content: “Hello!” }]
})
});
| Parameter | Type | Description |
|---|---|---|
| modelrequired | string | Model ID to use for inference. |
| messagesrequired | array | Array of message objects with role and content. |
| max_tokens | integer | Maximum tokens to generate. Defaults to 1024. |
| temperature | float | Sampling temperature, 0–1. Default 1.0. |
| system | string | System prompt injected before the conversation. |
| stream | boolean | Enable server-sent events streaming. |
GET
/models
List available models
Models
const models = await fetch(“https://api.aiexport.io/v2/models”, {
headers: { “X-API-Key”: process.env.API_KEY }
}).then(r => r.json());
| Parameter | Type | Description |
|---|---|---|
| family | string | Filter by model family: opus, sonnet, haiku. |
| limit | integer | Page size, max 100. Default 20. |
POST
/messages/stream
Streaming inference (SSE)
Streaming
const res = await fetch(“https://api.aiexport.io/v2/messages/stream”, {
method: “POST”,
headers: { “X-API-Key”: process.env.API_KEY },
body: JSON.stringify({ model: “apex-opus-3-20250401”, stream: true, … })
});
const reader = res.body.getReader();
while (true) {
const { done, value } = await reader.read();
if (done) break;
// Decode and process SSE chunks
console.log(new TextDecoder().decode(value));
}
POST
/batch
Create async batch job
Batch
const batch = await fetch(“https://api.aiexport.io/v2/batch”, {
method: “POST”,
headers: { “X-API-Key”: process.env.API_KEY },
body: JSON.stringify({
requests: [
{ model: “apex-haiku-4-20250401”, messages: […] },
{ model: “apex-haiku-4-20250401”, messages: […] }
]
})
});
| Parameter | Type | Description |
|---|---|---|
| requestsrequired | array | Up to 10,000 individual message requests. |
| webhook_url | string | POST notification URL when the batch completes. |
DEL
/batch/{batch_id}/cancel
Cancel a running batch
Batch
await fetch(`https://api.aiexport.io/v2/batch/${batchId}/cancel`, {
method: “DELETE”,
headers: { “X-API-Key”: process.env.API_KEY }
});
04
Rate limits
Free
Requests / min60
Tokens / day100K
Batch size100
ModelsHaiku only
Pro
Requests / min600
Tokens / day10M
Batch size5,000
ModelsAll
Enterprise
Requests / minCustom
Tokens / dayUnlimited
Batch size10,000
ModelsAll + Fine-tune
05
Changelog
Apr 2025
Apex Sonnet 4 released
2x faster inference, improved code generation, and extended 200K context window. Now the default recommended model for production workloads.
Mar 2025
Batch API goes GA
Async batch processing now generally available with up to 10,000 requests per job, webhook notifications, and 50% cost savings versus synchronous.
Feb 2025
Structured outputs (JSON mode)
New
response_format: { type: "json_object" } parameter guarantees valid JSON output on all Sonnet and Opus models.Jan 2025
API v2.0 launch
Unified endpoint for all model families, revised authentication with scoped API keys, and new token-usage response fields.
06
SDKs
Node.js / TypeScript
npm install @aiexport/sdk
Python
pip install aiexport
Go
go get aiexport.io/sdk
Bestseller #2
Bestseller #3
Bestseller #4
Bestseller #5