Docs Limits
Limits
Logo Lava is free to use. These are the limits that keep it that way.
B Klug, updated October 25, 2026
The service runs on one small server, so every request is counted and some are held back. The numbers below are the current ones. Requests without a key are counted for the address they come from, and requests with a key are counted for the key.
Request limits
| No key | Free API key | |
|---|---|---|
| Logos per minute | 60 | 300 |
| Logos per day | 1,500 | 50,000 |
| New AI briefs per day | 20 | 300 |
| Wait between new AI briefs | 30 seconds | 2 seconds |
Going over a limit returns HTTP 429 with a Retry-After header that tells you how many seconds to wait. Image requests get an image back that explains it, and JSON requests get a JSON error. An AI request that is only over the brief limit still returns a logo, drawn by the standard engine; see AI mode.
Cached images are free
An image that was drawn before is served from a cache and does not count against any limit. A request that includes a seed and the same options always names the same image, so after the first request it is cached on the server and, because the response is marked immutable for a year, usually in the visitor's browser and at the CDN as well. A site that shows a fixed set of logos therefore uses almost none of its allowance. Requests without a seed are never cached, since each one is a new logo, and each of them counts.
Limits of one request
| Item | Limit |
|---|---|
text | 60 characters |
| Whole URL | 6,000 characters |
| Width and height | 16 to 2048 pixels each |
scale | 1, 1.5, 2, 3 or 4. Other values snap to the nearest of these. |
| Output size | About 16.8 million pixels. Above that, the scale is lowered and a warning is sent. |
| Variations per call | 48 |
| Render time | 5 seconds |
| Renders at the same moment | 2, then a queue of 40. After that the answer is 503 with Retry-After. |
The server draws two logos at once and keeps a short queue behind them. During a rush you may see a 503. Waiting for the number of seconds in Retry-After and trying again is the right response. Keep your own concurrency low, a few requests at a time, when you fetch many logos in a loop.
Using it in a product
The terms allow you to use the logos Logo Lava makes, including in commercial work. If you expect steady traffic, use a key, pin seeds so the images cache, and download and store the logos you rely on rather than fetching them on every page view. If your needs are larger than the limits above, write through the contact page.
Next: API keys