Maestro
One-time purchase of the app. Provider credits are not included.
Meet Maestro
Maestro license and AI usage
Maestro desktop will be sold as a one-time purchase, with no app subscription. You connect your own AI accounts and pay providers directly for usage, according to each service’s rates and terms.
AI provider
Billed for API usage, according to the provider’s rates and terms.
Your choice
Models, modules, and features used for each task.
What the purchase includes
The purchase is for a Maestro license. AI credits, balances, and any included usage allowances belong to your provider accounts and are not included in the app price. Check the sales terms when the product becomes available.
Your account, your key
Create an API key in the account of the provider you want to use, then configure it in Maestro. This approach is known as BYOK, short for “bring your own key.” Configure only the providers you need; the app does not sell keys or handle their billing.
A chatbot subscription and an API are different services
Having a chatbot subscription does not mean you have a balance to use its API in Maestro. Check with the provider how to enable API access, which rates apply, and what usage limits you can set.
Costs depend on the work
The model, the amount of input and output text, the modules selected, and features such as web search affect usage. Combining more analyses can increase the cost. Batch processing and other modes have their own terms when available.
Prices updated automatically
With automatic updates enabled, Maestro checks provider prices daily while the app is open. In Settings, open View Prices to see the table and update date, or refresh it manually. If the check fails, the app keeps the last valid catalog; updating prices does not add new models.
Track estimates
Maestro uses available prices to estimate new tasks and records costs in the history. Updates do not recalculate costs already saved. The amount actually charged is the amount reported by the provider, which may include taxes, exchange rates, and other terms. Also check your account dashboard.
How to spend less with Maestro
For work that can wait, lightweight OpenAI models with Batch processing are a good starting point. Try them first: the lower-cost option still needs to deliver the quality your text requires.
Choose the model and processing mode
In the Advanced profile, you can choose each module’s model by name. Look for a lightweight OpenAI model that supports batch processing. In New submission, select Batch at the Confirm step. This mode is also available in the Basic profile.
Understand the discount and timing
OpenAI offers a 50% discount on eligible calls sent through the Batch API compared with standard direct processing. You can use Batch with a single document. The window is up to 24 hours per batch. Work with several stages can take longer: this window does not cover the entire review.
Use the modules you need
Under Modules in this submission, uncheck Use in this submission for modules that do not help with the task. Each AI module adds processing; choose those that serve the review’s purpose.
Start with a small sample
First review a short, representative document. Check the suggestions and cost before submitting more files. If the quality falls short, adjust the modules or try a more capable model only where needed.
Check what runs outside the batch
Web search and URL lookups use direct processing even when you select Batch. Final review and other direct calls also add costs without the batch discount. Check these options and whether direct processing is enabled if a batch fails.
Compare the estimate with the bill
Maestro’s Estimated cost is an approximation based on tokens, not the provider’s bill. Separate tool fees, taxes, and currency conversion may be excluded. In the app, open Settings → View prices… and check OpenAI’s current prices. Monitor usage and billing in your provider account.
OpenAI Batch API documentation
OpenAI terms checked on October 16, 2026.
Using the app
Open Maestro Help to review the steps and controls in your version. File and web search availability depends on the model and mode you choose.
Frequently asked questions