What Thumblore actually does
Thumblore turns a description of your video into finished thumbnail images — 1280×720 for YouTube, or 9:16 for Shorts, Reels and Stories. It is not a canvas editor with AI bolted on. You describe the video, pick the face you want in it, and it composes the whole frame: subject, background, lighting, headline text.
Then you edit what came back. Every generated thumbnail opens in an editor where you can change the mood, swap the face, rewrite the text in place, move the person, or just type what you want changed in plain English. Nothing you do there requires design software.
Your first thumbnail, in five steps
New accounts start with 100 free credits and no card on file. That is enough to make several thumbnails and edit them before you decide anything.
- Sign up and open the dashboard. Go to app.thumblore.com and sign in. The dashboard is the generation screen — it is the first thing you land on.
- Save a face first. Before generating, go to Avatars and upload one clear photo of yourself, or design a character if your channel is faceless-but-branded. Saving the face once is what makes every future thumbnail look like the same channel.
- Fill in the brief. Back on the dashboard: name the template, pick a scene mood, choose your avatar, and type the headline you want on the image. Add a one-line scene description if the video has a specific setting.
- Generate. One generation returns two variations so you have something to choose between. It usually takes under a minute.
- Edit and download. Open either variation in the editor, fix whatever is off, then download. Free exports carry a small Thumblore badge; any paid plan removes it.
Do this before anything else
Upload your face once, on day one. Almost every quality complaint about AI thumbnails — "it does not look like me", "every video looks like a different channel" — is the result of generating without a saved avatar. One upload fixes both.
What a good brief looks like
The generator reads your inputs as a single instruction, so vagueness in equals vagueness out. Compare these two briefs for the same video:
- Weak: Template name "video 12", no mood, no scene description, headline "MY NEW SETUP".
- Strong: Template name "Desk setup tour — 2026", mood "Bright & Fun", scene description "a clean white desk at night with RGB backlighting and three monitors", headline "MY $8,000 SETUP", reaction "Shocked".
The second one gives the model a subject, a setting, a light source and an emotion. It will get the frame right the first time far more often — which matters, because every regeneration spends credits.
Where everything lives
| Screen | What it is for |
|---|---|
| Dashboard | The generation brief. Format, mood, avatar, text, references. |
| Projects | Your thumbnails grouped by project and template name. See collections. |
| Avatars | Saved faces and characters. Upload, generate, delete. |
| Templates | Proven layouts you can set as a style reference. |
| Downloads | Every thumbnail you have generated, in one grid. |
| Billing & Credits | Plan, credit balance, top-ups. |
| Settings | Display name and the watermark switch. |
Where to go next
If you are setting up a channel rather than making one thumbnail, read the channel consistency guide — it covers the small number of decisions (one face, one palette, one text treatment) that make a row of your videos read as one channel. If you are rebuilding an existing channel's back catalogue, start with the channel makeover guide instead.