A Fortune 500 technology company ran a 50,000-video campaign on this API in three weeks. Now it is an endpoint in your stack: send a document or a prompt, get a finished, on-brand MP4 back. Your code can call it. So can your agent.
Most video APIs return a clip you still have to produce around. GLX returns the produced asset: script, voiceover, b-roll, brand treatment, and localization handled in one async job. You send source content and get back an MP4 you can put in front of a customer.
A render is a single job, not a toolchain you assemble. Story Builder writes the narrative from your content, then the platform produces the video around it and hands you a hosted MP4.
The product runs on the same API you integrate, and the platform is MCP-native. Exposing it to agents was a design decision, not a retrofit, so agentic callers get the same auth and controls your application code gets.
Start with a single render, scale to a campaign, or hand the keys to an agent. Same auth, same templates, same credits across all three.
Post a PDF, a presentation, or a prompt and get one finished video back. This is the fit for product flows that turn a document into a watchable asset on demand: a proposal, a report, a release note.
Read the docs →Feed a CSV, a form, or API calls and render one personalized video per row, recipient, or account. This is the motion behind 50,000 videos in three weeks, and it runs the same at 500.
Read the docs →Connect Claude or any MCP client and video generation becomes a tool call. An agent can research a prospect, brief the render, and deliver the finished personalized video, hands-free.
Read the docs →Every render, whether triggered by a person, a pipeline, or an agent, follows the same shape. Create a context, choose a template, collect the result.
Upload the material the video should be built from: a PDF, a presentation, or a written prompt. GLX indexes it so the script is written from your content, not from a generic template library.
Templates carry the narrative structure, brand treatment, and typed variables. Pick one and pass the specifics for this render: account name, product line, language.
The render runs as an async job. Register a webhook to be notified when the MP4 is ready, or poll the job until it completes, then deliver the hosted video wherever it needs to go.
# 1. Create a context from your source content ctx = glx.context.create(file="q3-launch-deck.pdf") # 2. Choose a template, pass your variables job = glx.storybuilder.render( template = "customer-outreach", context = ctx.id, variables = { account: "Acme Corp" }, language = "de" ) # 3. Webhook fires (or poll) when the render completes on_complete(job) # => finished MP4, hosted and ready to deliver
Pseudo-flow for orientation. Exact endpoints, auth, and schemas live in the developer docs.
You make one call. The platform handles what a production team would, on every job, every time.
Turns a PDF, a PPT, or a prompt into a structured narrative before a single frame renders. The story is the engine; the video is the output.
Voiceover in 75+ languages through ElevenLabs. Override language per render and ship localized variants from one template, no rebuild.
Licensed footage from the Storyblocks library, matched to the script automatically. No footage hunt, no licensing detour.
Bumpers, watermarks, colors, and logo treatment applied on every render. An agent-triggered video comes out as on-brand as a human-triggered one.
Your uploaded footage and images are auto-tagged, so renders pull the right asset by meaning, not by filename.
Fire the job and move on. A callback tells your app or agent when the video is ready, so nothing sits around polling.
Four patterns cover most production deployments. Each one composes the same three-step flow a different way.
Add a "watch this instead" button to anything your product outputs: proposals, reports, summaries. One render call per document.
One video per recipient, driven from your CRM, a CSV, or your warehouse. Campaigns handles the fan-out; you handle the data.
A signup, a renewal date, or a milestone fires a render, and the webhook drops the finished video into the next touchpoint.
An agent reads the meeting notes, renders a recap video in the customer's language, and queues it for review before send.
A finished video is only useful where your users are. GLX connects to the systems that carry it the last mile, so the render lands as a CRM asset, a CMS entry, or a file in the right folder.
Trigger renders from CRM records, publish to your video CMS, or drop files into the stores your teams already open. Full list and setup guides in the developer docs.
Built to clear your security review as fast as it clears your code review, with a pricing model finance can forecast.
Audited controls across the platform, with documentation available for your security review.
Single sign-on against your identity provider, so access follows the policies you already enforce.
Data handling aligned to GDPR, so global deployments do not stall in legal review.
1 credit = 1 video. No per-minute meters, no render-length surprises, and unit cost drops below $1 at enterprise scale.
Start free on Starter, then scale through Teams at $149 and Business at $499 to an Enterprise agreement with volume pricing. See pricing →
MCP support is not an adapter bolted onto the API. The server is native to the platform, so an agent gets typed tools with the same auth, rate limits, and audit trail your application code gets. Video becomes something an agent can produce, not just link to.
A concrete flow: an SDR agent researches a prospect, writes a one-paragraph brief, and calls GLX to render a personalized video in the prospect's language. The webhook returns the finished MP4 and the agent queues it into the sequence in Outreach. Nobody opened an editor.
If you are building copilots, internal automations, or AI-native sales and learning workflows, the MCP server is documented alongside the REST API. And if you want to pressure-test an architecture before you write a line, we will put an engineer on the call, not a rep.
Talk to an engineer →Pricing is credit-based and predictable: 1 credit = 1 video, and unit cost drops below $1 at enterprise scale. Plans run from a free Starter tier through Teams at $149 and Business at $499 to Enterprise agreements.
A PDF, a presentation, or a written prompt for single renders. Campaigns generates bulk personalized video from a CSV, a form, or API calls, one video per row, recipient, or account.
Through the native MCP server. Connect Claude or any MCP client and an agent can research a prospect, brief a render, and deliver the finished personalized video as typed tool calls, hands-free.
Yes. SOC 2 Type II, SAML SSO, and GDPR-aligned data handling, with documentation available for your security review.
Create an account, grab a key, and send your first document. The docs cover the rest, and an engineer is a message away when you want a second pair of eyes on the architecture.
Talk to an engineer →