You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
4 open-source models. OpenAI-compatible. $2 for 1,000 requests.
Quick Start
# Get a free trial (50 requests)
curl -X POST https://api.shiftai.dev/trial
# Generate text
curl -X POST https://api.shiftai.dev/v1/generate \
-H "Authorization: Bearer YOUR_KEY" \
-H "Content-Type: application/json" \
-d '{"prompt": "Explain quantum computing in one sentence", "model": "qwen2.5:3b"}'
SDK
npm install shift-ai
importShiftAIfrom'shift-ai';constai=newShiftAI('sk_your_key');constres=awaitai.generate('Hello world');console.log(res.response);// Chat (OpenAI-compatible)constchat=awaitai.chat([{role: 'user',content: 'Write a haiku about coding'}]);
CLI
npx shift-ai --key sk_xxx "What is 2+2?"
Models
Model
Size
Best For
qwen2.5:3b
1.9GB
Fast responses, simple tasks
gemma3:4b
3.3GB
Balanced speed/quality
qwen2.5:7b
4.7GB
Complex reasoning
gemma3:12b
8.1GB
Highest quality
Pricing
Plan
Requests
Price
Trial
50
Free
Starter
1,000
$2
Builder
5,000
$5
Pro
20,000
$15
Unlimited
100,000
$49
Why Shift AI?
Cheap: 10-50x cheaper than OpenAI for simple tasks
Fast: Models run on dedicated GPU, no cold starts
Private: No data logging, no training on your data
OpenAI-compatible: Drop-in replacement for /v1/chat/completions
About
Cheap, fast AI inference API. 4 open-source models. OpenAI-compatible. $2/1000 requests.