Model tiers
This doc describes the model tiers feature that allows you to choose between different models depending on your requirements
How to configure
You can head over to the workflow setup section of your model and then choose the model tier
_Note: This feature is available to all users on the credits accelerate plan, if you don't see this then contact support or your account manager for enabling i_t

You can now select between 3 model tiers
-
Nanonets Nano
Balanced speed and accuracy. Good for semi-structured docs with moderate variance eg: invoices, purchase orders, single page documents
-
Nanonets Mini
High accuracy with reasonable latency. Handles complex layouts and noisy scans well Use when: structure varies, or you require reliable table extraction. eg: bank statements, multi page invoices
-
Nanonets Pro
Best-in-class accuracy. Slowest and most expensive. Optimized for edge cases and mission-critical docs Use when: precision is non-negotiable—e.g., financial, forms, medical reports
Info and Limits
Feature | Nano | Mini | Pro |
---|---|---|---|
Credits per page | 0.3 | 0.5 | 0.8 |
VLM support | No | Yes | Yes |
Document level processing support | Yes | Yes | Yes |
Max file size limit in document level | 10 | 30 | 100 |
Data formatting support | No | Yes | Yes |
Updated about 13 hours ago