When you use OpenCRM's AI features, your usage is billed by your AI provider. You have a billing relationship directly with that provider, and their account is where you'll see what you've used and what you've been charged.

No. OpenCRM doesn't charge for AI usage and doesn't add a markup to your provider's costs. There's also no fixed fee for OpenCRM's AI features, so what you pay depends on how much your team uses them.

Most customers use an API account with their provider, which is pay-as-you-go. You add credits to your account in advance, and usage is deducted from that balance in near real time as your team uses OpenCRM's AI features. Once your credits are used up, you'll need to add more to keep using them.

Because credits are used as you go, it's worth keeping an eye on your balance. Many providers let you see your remaining credit in their account dashboard and set low-balance alerts, so you can top up before you run out.

Most providers also offer automatic top-ups, so you never run out. They'll notify you of any top-ups, so you can keep an eye on costs.

Each AI provider sets its own prices, and prices can differ between them. Even within the same provider, the price depends on the model you choose.

A helpful way to think about models is to compare them to members of a team. A junior colleague is quick and cost-effective for straightforward tasks, like pulling out a contact's details or tidying up a note. A skilled senior specialist costs more for their time, but they handle complex, nuanced work better, such as spotting patterns across a pipeline or preparing a detailed account summary ahead of a renewal.

AI models work in a similar way. Smaller, faster models cost less per token and suit routine tasks. Larger, more capable models cost more per token but are better at complex reasoning and more demanding work. Newer models can also change pricing, so it's worth checking your provider's pricing page for current rates.

Choosing the right model for each task is one of the most effective ways to manage costs. Your provider's documentation and your OpenCRM administrator can help you decide which model suits your team's needs.

AI providers measure usage in tokens. Tokens are small chunks of text used by AI models for processing and billing. As a rule of thumb, 1,000 tokens is roughly equivalent to 750 words of English text.

Each time the AI works on a request, you're charged for two things:

Many people using a standalone AI chat ask short questions and get short answers. OpenCRM works differently. When you ask the AI about a customer or your pipeline, it needs the relevant data to give a useful answer, so each request is typically larger than a standard chat message. This is expected, and it's what allows the AI to work with your real information.

Approximately. OpenCRM keeps an audit log that records approximate token counts for AI requests. These figures are for guidance only. They can help you understand patterns in your usage, but they aren't exact and don't match the charges on your provider's bill.