Subscription Plans - Begins w/ AI

Starter

$299/mo

10 requests/minute

~432K monthly capacity • No daily caps • Unlimited tokens

Supported models
- GPT-5.2
- GPT-5.2 Thinking
- Grok-4.1
- Grok-4.1 Thinking
- Gemini 3 Pro
- Gemini 2.5 Pro
- Gemini 2.5 Flash
- DeepSeek v3.2
- New models added automatically as they launch
Queue-based processing
- All requests are queued and processed asynchronously
- No live streaming supported
Unlimited tokens
- No token-based billing — only requests per minute count
- Each model respects its native token limits
Response delivery
- Polling endpoint available
Automatic retries
- Up to 3 retry attempts on failures
- Optional fallback model if primary is unavailable
Task completion time
- Lightweight models 15 – 45 seconds
- Standard models 30 – 90 seconds
- Thinking models 2 – 5 minutes
Response retention
- Requests and responses stored for 7 days for review and debugging
3 API keys
Email support (48-hour response)

Scale

$699/mo

30 requests/minute

~1.3M monthly capacity • No daily caps • Unlimited tokens

Supported models
- GPT-5.2
- GPT-5.2 Thinking
- Grok-4.1
- Grok-4.1 Thinking
- Gemini 3 Pro
- Gemini 2.5 Pro
- Gemini 2.5 Flash
- DeepSeek v3.2
- New models added automatically as they launch
Queue-based processing
- All requests queued and processed asynchronously
- No live streaming supported
Unlimited tokens
- No token-based billing — only requests per minute count
- Each model respects its native token limits
Response delivery
- Polling endpoint available
Automatic retries
- Up to 3 retry attempts on failures
- Optional fallback model if primary is unavailable
Task completion time
- Lightweight models 10 – 30 seconds
- Standard models 20 – 60 seconds
- Thinking models 1 – 3 minutes
Response retention
- Requests and responses stored for 30 days for review and debugging
10 API keys
Priority email support (24-hour response)

Enterprise

$1,699/mo

100 requests/minute

~4.3M monthly capacity • No daily caps • Unlimited tokens

Supported models
- GPT-5.2
- GPT-5.2 Thinking
- Grok-4.1
- Grok-4.1 Thinking
- Gemini 3 Pro
- Gemini 2.5 Pro
- Gemini 2.5 Flash
- DeepSeek v3.2
- New models added automatically as they launch
Queue-based processing
- All requests queued and processed asynchronously
- No live streaming supported
Unlimited tokens
- No token-based billing — only requests per minute count
- Each model respects its native token limits
Response delivery
- Polling endpoint available
Automatic retries
- Up to 3 retry attempts on failures
- Optional fallback model if primary is unavailable
Task completion time
- Lightweight models 5 – 20 seconds
- Standard models 10 – 40 seconds
- Thinking models 45 – 90 seconds
Response retention
- Requests and responses stored for 90 days for review and debugging
Dedicated infrastructure allocation
- Reserved worker capacity for your requests
- Isolated queue processing
Unlimited API keys
Priority email support (12-hour response)

Choose Your Plan