Simple, Consumption-Based Pricing

Only pay for what you use, with transparent pricing

Model Pricing Details

Platform Fee

After free credits are used

$0.50

per 100 predictions

Model Usage (at cost)
Claude 3.5 Sonnet
Input tokens$0.003/1K
Output tokens$0.015/1K
Claude 3.5 Haiku
Input tokens$0.001/1K
Output tokens$0.005/1K
✨Eligible startups, and open source maintainers get $240 worth of free credits over 6 monthsApply →

Free

Perfect for getting started

$0
per month
  • Free forever
  • $5 free model credit
  • 500 free predictions
  • Up to 100 functions
  • 2 clusters

Pay as you go

For growing projects

$0.50
per 100 predictions
  • Everything in free tier
  • Up to 20 clusters
  • Up to 2000 functions (soft limit)
  • Higher rate limits
  • Priority email support

Enterprise

Maximum control and security

Custom pricing
per year
  • Deploy on your infrastructure
  • Bring your own model keys
  • Full data isolation
  • Dedicated Slack support
  • Custom SLAs

Understanding Your Costs

Simple, transparent pricing based on usage

Total Cost = Prediction Costs + Model Costs

Both metered and billed through Inferable

Free Tier Includes

• First 500 predictions free
• $5 in model credits free

Prediction Costs

$0.50 per 100 predictions after free tier. This covers our platform costs for running the agent runtime and managing clusters in our cloud.

Model Costs

We provide access to Anthropic's Claude models at cost with no markup. Model usage is metered per prediction and billed at the exact rates charged by our model provider. You only pay for the model compute your agents actually use.

Frequently Asked Questions

Everything you need to know about pricing

Pricing