Connect a Google Gemini API Key to Tubeup
Create a free Google Gemini API key in Google AI Studio and add it to Tubeup so the app can write SEO titles, descriptions, tags, and hashtags for a whole batch.
On this page
Tubeup uses Google Gemini to write your video metadata, and it uses your own API key rather than reselling access through us. You create a key in Google AI Studio, paste it into Tubeup's settings, and from then on one action generates titles, descriptions, tags, and hashtags for an entire batch.
Before you begin
- Tubeup installed and activated
- A Google account
- A few minutes at aistudio.google.com
Why you supply your own key
Bringing your own key means the AI requests go from your machine to Google directly. Nothing routes through a Degird server, so there is no intermediary holding your video titles or descriptions, and there is no per-video credit system layered on top of the $19 one-time price.
It also means the usage terms and limits are the ones Google publishes for your key, not ones we invented. Gemini has a free tier that is comfortably enough for normal publishing volume. If you eventually exceed it, you deal with Google directly and Tubeup keeps working with the same key.
Create and add the key
Open Google AI Studio
Go to aistudio.google.com and sign in with the Google account you want the key billed and rate-limited under. This does not have to be the same account as your YouTube channel.
Create an API key
Find the API key section of AI Studio and create a new key. Google will show you the key string once; copy it immediately, because the full value is not always retrievable afterwards.
Paste the key into Tubeup
In Tubeup, open Settings and find the field for the Gemini API key. Paste the key in and save. Nothing else about the app's behaviour changes until you save.
Test on one video
Generate metadata for a single video rather than the whole batch. If a title, description, tags, and hashtags come back, the key is working. If nothing comes back, the key is usually the problem before anything else is.
Generate for the batch
Once a single video works, run generation across the entire batch in one action and move on to reviewing the output.
Treat the key like a password. Anyone who has it can spend against your Google quota. Do not paste it into a shared document, a screenshot, or a support ticket, and revoke it in AI Studio if you think it has leaked.
What Tubeup generates
With a key in place, generation covers the fields that actually affect discovery on YouTube:
| Field | Shape | What you get |
|---|---|---|
| Title | Single line | An SEO-oriented title drafted from the video, which you can edit before upload. |
| Description | Long form | A full description body, written to be readable rather than keyword-stuffed. |
| Tags | List | A set of keyword tags for the video. |
| Hashtags | List | Hashtags suitable for the description or title. |
| Social copy | Long form | Post copy for Facebook, X, LinkedIn, and Instagram, generated on demand after upload. |
Generation is a draft step, not a publish step. Nothing goes to YouTube until you review the results and start the upload yourself.
Reviewing what comes back
Read the titles first. Titles carry most of the click-through weight on YouTube, and they are the field where a human edit pays for itself fastest. Descriptions usually need less work, though it is worth checking that any claim in the text is one you would actually make about the video.
If the output feels generic across a batch, the usual cause is uninformative filenames. Videos named descriptively give the model something to work with; videos named after camera export defaults do not. Fixing this happens in your base folder setup, not in the AI settings.
Generate metadata before you set schedules and visibility, not after. Reviewing titles tends to change your mind about publishing order, and it is easier to assign publish dates once you know what each video is actually called.
If generation fails
Check the key first: confirm it was pasted in full with no trailing spaces and that it still exists in AI Studio. Then check your internet connection, since Gemini generation is one of the three network calls Tubeup makes. If you have been generating heavily, you may have hit a Google rate limit, in which case waiting and retrying is the fix rather than changing any Tubeup setting.
Next steps
Tubeup
Cross-platform desktop app that bulk-uploads videos to YouTube with AI-generated metadata, scheduling, and multi-channel support.
See also
Related documentation
- AI-Generated Titles, Descriptions and TagsHow Tubeup uses Google Gemini to write SEO titles, descriptions, tags and hashtags for a whole batch of videos, and how to review and shape the output.
- Getting Started With TubeupInstall Tubeup, connect a YouTube channel, and publish your first batch of videos with AI-generated titles, descriptions, and tags.
- Set Up Your Tubeup Base FolderChoose a base folder so Tubeup scans and registers your videos automatically, matches thumbnails by filename, and moves published files out of the way.