B-roll, media and what Valmera generates
Valmera edits footage you provide. It doesn’t generate whole videos, AI images or AI video clips from a prompt. Instead it finds and builds the supporting media an edit needs: stock b-roll and photos, real footage of a named topic, songs and sound effects from the web, recordings of a live website, and title cards, color screens, graphics and freeze frames rendered inside the edit.
What Valmera finds and builds
| You need | How Valmera handles it | Ask |
|---|---|---|
| Generic b-roll | Searches stock video and photo libraries, checks the frames, places the shot | “B-roll of a busy city when I mention the commute” |
| Footage of a named person, product or event | Finds real footage on the web and cuts to it at the mention | “Show the Starship launch when I mention SpaceX” |
| A photo of a real subject | Searches openly licensed photo records | “Show a photo of the Golden Gate Bridge” |
| Music | Your file, a link, or a named song found on the web | “Use ‘Blinding Lights’ by The Weeknd” |
| Sound effects | Real recordings from Openverse and Freesound | “A camera shutter at 0:12” |
| Footage of a website | Records a public page, or a scripted demo with a visible cursor | “Record valmera.io scrolling for 10 seconds” |
| Cards and interstitials | Renders title cards, color screens and glitch screens in the edit | “Cut to a black card that says ‘Chapter 2’” |
| A whole video from a text prompt | Not available: Valmera edits footage you provide | — |
| AI-generated images or clips | Not available: upload your own or use stock | — |
Wondering whether you need an editor or a generator? See AI video editor vs. AI video generator.
Add stock b-roll
- Say what should be on screen, and when
For example: “Add b-roll of a busy city street when I mention the commute.” Concrete, visible subjects find better shots than moods.
- Valmera searches stock libraries
It searches stock video and photo libraries, matched to your video’s orientation. For a whole video it plans b-roll as a sequence and can decide a moment is better without any.
- It checks the actual clip
Valmera downloads the chosen clip, looks at its real frames and picks the right seconds before placing it.
- It places the shot
Usually as a full-frame cutaway while your voice keeps playing, or as an insert. The reply names the shot it used.
- Review and swap
Ask for a different shot, a shorter cutaway, or remove it.
- Add b-roll of a busy city street when I mention the commute.
- Plan b-roll for this whole video, only where it helps the story.
- Use a vertical clip of ocean waves under the intro, and keep my voice.
- Swap the coffee shot at 0:42 for something brighter.
Type these in Studio chat, or ask Claude, ChatGPT, Cursor or another assistant connected to Valmera’s MCP server. More in the prompt library.
Stock clips are silent. A cutaway plays over your footage while your voice continues; an insert pauses your footage and adds time. See overlays versus inserts.
Real footage of a named topic
When the speaker mentions a specific person, company, product or event, Valmera can search the web for real footage of it, download a short clip, pick the right seconds and cut to it at the mention. It tells you which video it used (title and channel). Make sure you have the right to use that footage.
Import from a link
Paste a direct file link or a YouTube, TikTok, Vimeo or SoundCloud page and say how to use it. Link imports take video up to 500 MB, audio up to 50 MB and images up to 10 MB, up to 60 minutes long, with video at up to 1080p. Valmera reports the source of each import; downloading media doesn’t grant the right to republish it. Full details: link imports.
Record a website or product demo
| Mode | What it records | Ask |
|---|---|---|
| Scroll capture | A public page at your video’s shape: holds the top, scrolls smoothly to the bottom, holds. 4 to 30 seconds (12 by default), silent. | “Record valmera.io scrolling for 10 seconds” |
| Scripted demo | A browser with a visible cursor clicks, types, hovers and scrolls through the steps you describe, then Valmera places it with a zoom on each click. | “Record a demo: click Pricing, scroll to the plans, hover over the yearly toggle” |
Four to ten steps that tell one story make the best demo. Recordings show the public site as a visitor sees it: Valmera can’t log in and won’t type into password or payment fields.
For any screen recording, including ones you made yourself, Valmera can also make a small or jittery cursor bigger and steadier with click ripples, and float the recording in a rounded window on a color or gradient background. See the website capture example.
Title cards, color screens and glitch screens
| Element | What it is | Ask |
|---|---|---|
| Title card | A full-frame card with your text, then back to the footage (2 to 3 seconds reads well) | “Cut to a black card that says ‘Chapter 2’ at 1:30” |
| Color screen | A solid or two-color gradient interstitial | “A white flash screen at 0:58 for half a second” |
| Glitch screen | A short digital, VHS or TV-static glitch between sections, with an optional static burst | “A half-second glitch between the podcast and the demo” |
| Freeze frame | A held still with big text over it | “Freeze at 1:12 with ‘The moment it worked’” |
These render inside the edit, so there’s nothing to download or generate. Cards and screens add their length to the video; a freeze frame does too unless you ask to keep the audio running.
Graphics and layouts
- Vector graphics: arrows, rings, underlines, highlight panels and progress bars that point at things in the frame.
- Editorial graphics: statements, comparisons, metrics, quotes, chapter markers, labels and speaker headlines.
- Multi-panel walls: two or three clips playing side by side on black, up to 20 seconds.
- Picture cards: your footage inside a rounded card with an entrance and exit.
More on text and graphics: text, logos and overlays.
Slideshows from photos
Upload images or clips and ask for a slideshow or montage. Valmera builds a sequence from your files, gives each still a slow camera move (zoom in, zoom out or a pan) and a duration (three seconds per image by default), and can add your music and on-screen text.
- Make a 30-second slideshow from these photos with slow zooms and my music.
- Use these product shots as a montage, two seconds each, in the order I uploaded them.
Type these in Studio chat, or ask Claude, ChatGPT, Cursor or another assistant connected to Valmera’s MCP server. More in the prompt library.
Troubleshooting
| Symptom | Fix |
|---|---|
| You asked for a generated image or clip | Valmera doesn’t generate them. Upload your own, or ask for a stock photo or clip of the subject. |
| The b-roll doesn’t match | Describe a concrete, visible shot (“a barista pouring latte art”) and ask for another candidate. |
| No footage found for a named topic | Name the exact event or product, or paste a link to the footage you want. |
| The website recording shows a cookie banner or login page | Recordings show the public page as a visitor sees it. Use a page that loads without a login. |
| A link import failed | The link may be private, removed or blocked. Download the file and upload it instead. |
Limits
- No text-to-video generation of a whole video.
- No AI image or AI video clip generation, no AI music and no AI voices.
- Stock clips are silent; check each item’s license before publishing.
- Website recordings cover public pages only: no logins, and no typing into password or payment fields.
Tools behind this page: search_stock research_broll add_stock_media find_footage fetch_url record_website record_website_demo showcase_demo enhance_cursor set_screen_frame add_title_card add_color_screen add_corrupt_screen add_freeze_frame add_vector_graphic set_editorial_graphic compose_panels set_picture_card start_media_sequence insert_media add_overlay · Full toolset reference
Frequently asked questions
Can Valmera generate a video from a text prompt?
No. Valmera is not a text-to-video generator. It edits footage you provide, and finds or builds the supporting media an edit needs: stock b-roll, real footage of named topics, songs, sound effects, website recordings, title cards and graphics.
Can Valmera generate AI images or AI video clips?
No. Valmera doesn't generate images or video clips with AI. Upload your own images and clips, or ask for stock footage and photos, which Valmera searches, checks and places for you.
Where does Valmera's b-roll come from?
From stock video and photo libraries such as Pexels, Pixabay and Openverse (including openly licensed Wikimedia and Flickr photos), from real footage of named topics found on the web, and from clips you upload or link.
Is stock b-roll free to use commercially?
Licenses vary by source and item. Valmera reports each item's source and passes on license lines that carry an obligation, such as required credit. Check the terms before you publish.
Can Valmera record my website?
Yes. It can record a public page scrolling top to bottom, or run a scripted demo in which a visible cursor clicks, types and scrolls through the steps you describe. It can't log in and won't type into password or payment fields.
Can I import a clip from YouTube?
Yes. Paste the link and say how to use it. Link imports take video up to 500 MB and 60 minutes at up to 1080p. Use only media you have the rights to publish.