AI Studio feature reference.

Dashboard

The AI Studio command center for campaigns, templates, support resources, product orders, onboarding, and quick launch actions.

Campaign list

Shows existing campaigns, their current state, and the path back into generated results or campaign setup.

Template list

Shows reusable video templates, lets users open a template editor, and starts new campaigns from an existing template.

Support side panels

Surface guides, tours, office hours, quick links, and resource links without interrupting the main workflow.

Welcome template tutorial

Guides new users through creating a first variable, generating a sample, and watching the result.

A template starts with one source video. The current AI Studio create flow exposes five source paths.

Record a video

Use the browser camera and microphone in AI Studio. The recorder keeps script notes beside the capture experience.

Upload a video

Attach an existing recording. Camera and microphone access do not open, and AI Studio prepares the preview and transcript after upload.

Speakeasy video

Import a finished Speakeasy presenter video into AI Studio so it can be personalized like any other template.

Community template

Browse reusable starter templates, including the welcome template and category-filtered template store entries.

Text-to-video

Generate presenter videos from script text and use text-to-video templates when that source path is enabled.

Template name

Rename the video instance and keep the asset recognizable before launching campaigns.

Video preview

Review the source video, verify the uploaded asset is available, and check whether the recording can support Mini or Max presentation modes.

Transcript editor

Transcribes the source video, caches transcript data, lets users select words, creates variables, replaces duplicate variable definitions, and deletes variable spans cleanly.

Variables panel

Shows variable segments, transcript linkage, standard or premium voice status, and jump-to-transcript actions.

Voice settings

Standard voice is the default. Premium voice and tuning are available per variable for eligible plans.

Generation mode

Best quality uses the current full-script generation path. Preserve source video keeps variable lip sync behavior for templates that need it.

Manual timing

For no-lipsync workflows, variable timing windows store start time, window end time, anchor, and fit behavior.

Training audio

Advanced template controls support uploading training audio when the voice needs more source material.

Try sample

Generates one test video from sample variable values before a campaign is launched.

Create new campaign

Opens campaign setup after the template has a source video and at least one valid variable.

CSV and Excel

Upload spreadsheet data, normalize it for campaign use, preview rows, and map headers to template variables.

Zapier

Open the BHuman Zapier flow and send app-triggered data into a campaign.

Leadr

Use a selected AI Studio template inside Leadr outreach and LinkedIn-style campaign workflows.

Pabbly and API

Backend integration paths support object-style payloads, ordered arrays, callback URLs, and generated asset polling.

Variable mapping

Maps each template variable to a source column, ignores reserved output columns, and auto-selects exact header matches when possible.

Mini or Max mode

Campaigns choose a viewing mode before rendering. Mini requires background coverage around personalized moments; Max is plan and face-distance gated.

Manual audio recording

Lets users record custom variable audio instead of using the generated voice path, except on store templates where the control is disabled.

Do It For Me

Offers a managed handoff path when a user wants BHuman to help launch the finished campaign.

• Background source types: generic desktop, direct image or video link, public website screenshot, LinkedIn profile screenshot, and uploaded image or video.
• Presenter display modes: square, circle, and greenscreen/chroma key.
• Timing options: exact start and end time or show throughout the entire video.
• Validation catches invalid time ranges, missing source columns, missing uploads, unsupported direct media links, LinkedIn screenshot failures, and overlapping segments.
• Subtitles use the edited template transcript with row-specific variables inserted for each generated video.

Live status refresh

Generated results load over SSE and refresh every five seconds while active videos are queued, preparing, or processing.

Ready outputs

Succeeded or completed videos expose download, hosted share link, thumbnail, GIF preview, source row index, and row values.

Failed outputs

Failed cards show error messages and pipeline execution names when available.

Stale outputs

Videos that stop updating for long enough are marked as needing review instead of looking permanently in progress.

Retry generation

When the full batch failed and no active or ready videos remain, the results page can prepare the campaign for another generation attempt.

CSV and XLSX export

Download updated campaign rows with generated asset columns after rendering.

Email

Send generated videos from connected mailbox workflows, track sending progress, and open email reports when activity exists.

SMS

SMS delivery is represented in the share UI, with availability controlled while improvements are in progress.

Landing pages

Open the landing page builder from generated results to pair videos with CTA widgets and page content.

API

Open the public API reference from the share page when a workflow needs to consume generated assets programmatically.

Leadr handoff

Move into Leadr when the generated videos should be used in LinkedIn-style outreach.

Pipeline endpoints

Campaign generation, Zapier, Pabbly, Leadr, sample generation, preview generation, and pipeline webhooks are handled by AI Studio backend routes.

Generated video endpoints

Generated videos can be fetched by campaign ID, video instance ID, generated video ID, or user-scoped admin views where available.

Webhook settings

Saved webhooks and per-request callback URLs receive asynchronous generated asset payloads.

Swagger

The AI Studio backend includes Swagger UI pointed at openapi.json. Use it for endpoint-level schemas when the public docs summarize behavior.