Fair Use Policy

Last updated: 26 May 2026 · Version 1.0

Koda's paid plans say "unlimited" for AI-powered features like captions and brand intelligence scans. We mean it for any normal workflow — but because each of those features costs us money to run, this policy spells out where "unlimited" stops and "abuse" starts.

The short version: use the AI features as much as you need to run your business. If you write hundreds of captions a month for your clients, you'll never hit a limit. The caps below only kick in if usage looks like scripted, automated, or non-human activity.

What "unlimited" means

"Unlimited" on Koda means: there's no per-feature usage cap on you as a normal user using Koda the way it's designed to be used. There isn't a meter ticking down. We don't bill overage.

However, every AI request to Koda costs us API credits with our AI provider, Anthropic. To keep paid plans sustainable for everyone, we apply a soft fair-use threshold per plan. The vast majority of customers — over 99% based on our internal usage data — never come close to these thresholds.

Our fair-use thresholds (per month)

Plan	AI captions	Brand intelligence scans
Free	10 / month (hard cap)	2 total (hard cap)
Freelancer	500 / month	50 / month
Studio	2,000 / month	200 / month
Agency	10,000 / month	1,000 / month

For context, a Studio user managing 10 clients posting 5 times per week — with 2 to 3 regenerations per post — typically uses around 400–500 captions per month. The Studio threshold is roughly 4× that.

What happens if you hit the threshold

At 80% of the threshold we'll show a notice in the app and email the org admin.
At 100% we'll pause new AI requests until the next billing cycle resets the counter, OR you can upgrade to a higher plan at any point.
We will never charge you per-request overage fees. The only consequence of hitting the threshold is a pause on AI requests.

What counts as abuse

Use that goes beyond the fair-use thresholds and shows any of the following patterns is treated as abuse:

Scripted, automated or programmatic calls to the AI endpoints from outside the Koda app interface
Sharing an account between multiple agencies or unrelated organisations
Reselling Koda's AI output as a standalone product
Generating content for purposes prohibited by Anthropic's usage policies (e.g. illegal content, harassment, mass-produced disinformation)

If we detect abuse we'll contact the org admin first. Continued abuse may result in account suspension under our Terms of Service.

What's NOT abuse

Generating 50, 100, even 300 captions in a single day because you're catching up on a backlog
Re-running the brand scan on a client whose website you just updated
Heavy usage during shoot weeks followed by quiet periods
Building captions across all your clients at once

This is what Koda is for. Use it.

Changes to this policy

If we change the thresholds, we'll email all active customers at least 30 days before the change takes effect. We commit not to lower a threshold without that notice.

Questions

If you're hitting a limit and don't think you should be, or if you have an unusual high-volume use case, email [email protected]. We'd much rather have a conversation than enforce a cap.