Content Generation Lab

The Content Generation Lab is the media studio in Convia Pro. Use it to generate images, voiceovers and talking-avatar videos for your show, and keep every result in one asset library. Open it from In Motion → Content Lab in the launcher.

The Lab is plan-gated, so it appears under In Motion in the launcher only if your plan includes it. Viewers cannot generate; contributors, managers and owners can.

What the Content Generation Lab can make

The Lab's Capabilities list also shows Video as Coming soon. The Connections panel shows whether ElevenLabs and HeyGen are connected, with a link to connect each one.

Image costs

If a generation fails, its credits are returned. Costs can change; the Generate button always shows the current cost. See AI credits.

Generate an image in the Image Studio

The Image Studio lives at In Motion → Content Lab → Images. It has four tabs: Episode Thumbnail, Infographic, Quote Card and Graphic.

Make an infographic, quote card or graphic

1. Open the Infographic, Quote Card or Graphic tab.
2. Write a Prompt: the title and key points for an infographic, the quote and attribution for a quote card, or a description of the graphic.
3. Optionally attach reference images and Brand Kit assets. Reference images must be files uploaded to your workspace.
4. Pick an Aspect ratio: 1:1, 4:5, 9:16 or 16:9.
5. Click Generate. The image appears under Preview and is saved to the asset library.

Brand Kit assets are used according to their type: a template sets the layout to follow, a logo is placed as a corner mark, a background fills the background and an overlay acts as a frame. The Brand Kit is plan-gated, so this option appears only where your plan includes it.

If the image cannot be made, the message says why. A prompt blocked by safety filters says so. If the image model replies with text instead of an image, the message asks you to rephrase the prompt.

Make a thumbnail in the Lab

Thumbnails in the Lab use the same styles as Thumbnail Studio and start on your workspace's default style. Thumbnail generation is plan-gated; without it, the tab says it is not on your plan.

1. Open the Episode Thumbnail tab.
2. Choose From episode and pick an Episode, then choose which guest and host photos to feature. Or choose Generic (upload), add Media to include (optional) and type Thumbnail text (optional).
3. Pick the Platform the thumbnail is for and a Style. Styles you saved for the workspace appear under Your styles.
4. Optionally add Brand Kit assets.
5. Click Generate (25 credits).
6. Use Download for a platform to get its exact size, such as 1280×720 for YouTube.

For an episode, earlier thumbnails appear under Previously generated for this episode. For the full set of thumbnail controls, such as drafting an idea or editing a result, see Thumbnails.

Generate a voiceover

Voiceovers run on your own ElevenLabs account. Connect it first under Integrations → ElevenLabs. See Connect AI accounts.

1. In the Lab, open Voice.
2. Type or paste the script into Text (up to 5,000 characters).
3. Pick a Voice from your ElevenLabs account.
4. Pick a Model:
   • Multilingual v2: high quality, 29 languages.
   • Flash v2.5: low latency, cost-efficient.
   • v3 (expressive): the most expressive, and supports delivery tags.
5. With v3, use Delivery tags to insert cues such as [laughs], [whispers], [sighs], [excited], [sarcastic] or [curious] at your cursor. You can also insert a timed pause.
6. Choose an Output format: MP3 at 128, 64 or 32 kbps.
7. Click Generate voice. Listen under Preview, then download it, or change a setting and click Regenerate. Each run is saved as its own asset.

Generate an avatar video

Avatar videos run on your own HeyGen account. Connect it first under Integrations → HeyGen.

1. In the Lab, open Avatar.
2. Under Avatar, choose HeyGen avatar and pick one, or choose Your photo and upload a PNG, JPEG or WebP image. A clear, front-facing photo works best, up to 5 MB.
3. Write the Script: what the avatar should say (up to 5,000 characters).
4. Pick a Voice: a HeyGen voice, or an ElevenLabs voice if ElevenLabs is connected too.
5. Choose an Aspect ratio (9:16 vertical, 16:9 landscape or 1:1 square) and a Resolution (720p or 1080p). Optionally add a Title.
6. Click Generate avatar. Rendering usually takes 1 to 5 minutes. You can keep working: the video saves to the asset library when it finishes.

If you close the window during rendering, open the Lab again later. The asset library checks on any avatar still rendering when it loads.

Use the asset library

Everything you generate lands in the Asset library on the Lab's main page. Filter it by All types, Voice, Image, Video or Avatar, then preview, download or delete any asset. Lab assets are private to your workspace.

Generated images can be attached to posts, so Lab work flows into publishing.

Generate a visual from the post composer

You do not have to leave a post to make its image. Where your plan includes the Lab, the post composer has a Generate a visual with AI panel:

1. Write the Prompt, or click Draft from post to write one from your caption and then edit it. For a campaign post that is waiting on a visual, the prompt starts from the campaign's brief.
2. Pick a Type (Graphic, Infographic or Quote card) and an Aspect ratio.
3. Click Generate. To refine, edit the prompt and click Regenerate. Each result is kept as a variant you can switch between.
4. Select the variant you want and click Use this. The image attaches to the post. Save the post to keep it.

Drafting a prompt with Draft from post is free. Generating the image costs the same as in the Lab, and the panel shows the cost before you click. The article editor has the same panel, where Draft from article writes the prompt and the image becomes the article's featured image.

Related

• Thumbnails: Thumbnail Studio, styles, ideas and platform sizes in depth.
• Connect AI accounts: connect ElevenLabs and HeyGen for voice and avatar.
• AI credits: what images cost and how credits refresh.
• How AI uses your brand: how your Brand Brain and styles shape what the Lab makes.
• Posts: attach generated images and publish them.