Credits and limits

Some requests from connected AI tools use credits from your organization's balance. Requests that only count things or look up your setup are free. Reading conversations, taking inbox actions, and replying cost credits.

What each request costs

What the AI tool does	Cost
Counting conversations or checking comment sentiment	Free
Looking up accounts, smart labels, teammates, automations, or organizations	Free
Reading or editing the knowledge base	Free
Creating, editing, or deleting smart labels	Free
Unhiding a comment, or liking and unliking it	Free
Listing or searching comments, DMs, or live chats	1 credit per 50 returned
Checking for new or changed conversations (any channel)	1 credit per 50 returned
Listing or searching email threads	1 credit per 10 returned
Opening a single conversation (any channel)	1 credit
Grouping comments or DMs into themes	1 credit per 25 analyzed
Grouping email threads into themes	1 credit per 5 analyzed
Applying a label, assigning, resolving, or adding a note	1 credit each
Saving a reply draft or sending a reply	1 credit each
Moderating comments with Brandwise AI	1 credit per comment reviewed
Hiding or deleting a comment	1 credit each
Bulk labeling, assigning, or resolving	1 credit per conversation changed
Saving reply drafts in bulk	1 credit per new draft

A few rules apply to all of these:

• Each request is rounded up to the next whole credit, so any request that returns something costs at least 1 credit.
• A request that returns nothing is free.
• You're only charged when a request succeeds.
• When a teammate sends a draft the AI saved, there's no extra charge.
• If the AI retries the exact same reply, it isn't charged or sent twice.
• Moderating skips comments that were already reviewed, at no charge, unless the AI asks to review them again.
• In a bulk request, conversations that fail aren't charged, and neither is a draft that was already saved.
• Checking for changes again with the same request in the same clock minute returns the same answer at no extra charge. A retry that lands in the next minute is checked, and charged, again.

Examples

• "Show me the 20 newest comments." 1 credit.
• "Go through my last 120 comments." If the AI reads them in pages of 50, 50, and 20, that's 3 credits.
• "Read the 25 newest support emails." 3 credits.
• "What are the top 5 themes in my DMs?" Grouping into 5 themes looks at up to 250 conversations, so up to 10 credits. The same request for email threads costs up to 50 credits.
• "Draft replies to these 10 comments." 10 credits for the drafts, plus whatever it costs to read the comments.
• "Moderate my 20 newest unreviewed comments and hide anything rude." 20 credits for the reviews, plus reading the comments. Moderation hides what it flags at no extra cost; hiding something it left up costs 1 more credit.
• An AI tool that checks for new messages every few minutes. Free whenever nothing changed. When 12 conversations changed since the last check, 1 credit.
• "Label these 30 DMs as Shipping." 30 credits, minus any that couldn't be labeled.

When you don't have enough credits

Before a request runs, Brandwise checks that your balance can cover the most it could cost. If it can't, the request is refused and nothing is charged. Asking for fewer results or fewer themes lowers that amount.

Moderating comments works differently: each comment is checked and charged as it's reviewed. If your balance runs out partway through, the comments already reviewed stay reviewed and charged, and the rest aren't reviewed or charged.

Requests are also refused while your credits are on hold, which usually means there's a payment problem to fix. Free requests keep working either way.

When a request is refused for credits, the AI tool is given a link to your organization's Settings > Billing page and asked to pass it on to you. There you can add credits, upgrade your plan, turn on auto-upgrade, or fix a payment problem. You need permission to manage billing to open that page. For more on adding credits, see Add more credits.

See your MCP usage

Go to Settings > Billing > Credit Usage. Credits used by connected AI tools show up under Agent Access, except comment moderation, which shows up under Comment Moderation together with your automatic moderation. You need permission to manage billing to see this page.

Rate limits

To keep things fair for everyone, Brandwise limits how many requests can be made in a short time. If an AI tool hits a limit, nothing is charged, the tool is told to wait and try again, and the request shows as Rate limited under Settings > MCP > Recent activity.