Gamma

Let AI agents turn a prompt or source content into polished Gamma decks, documents, webpages, and social posts. Agents create generations, poll them to completion, and return a shareable Gamma link plus optional PPTX, PDF, or PNG exports.

3 actions API key available

Gamma website ↗ Platform Documentation ↗ Capability data verified 2026-07-28

A new project brief arrives through the intake form

The agent assembles the generation input: the brief's content, the house theme name, tone, audience, and card count

Create Generation submits the deck request and returns a generation ID immediately

Get Generation polls in the background until status reads completed

The agent verifies the generation completed rather than failed and records the credits deducted

The account lead opens the Gamma link, reviews the deck's claims and framing, and approves or requests a regeneration with adjusted instructions

On approval, the shareable link and PPTX export post to the team channel and archive to the shared drive

What This Integration Enables

Most content a business presents already exists somewhere as prose before someone spends a night converting it into slides. Gamma collapses that conversion, and this connector puts it inside your workflows: agents feed a prompt or source content in, choose presentation, document, webpage, or social output, control theme, tone, card count, and image sourcing, and get back a shareable Gamma link plus an optional PPTX, PDF, or PNG export. Generation is asynchronous, so flows either fire-and-poll or block on a single call when latency does not matter. The position: generation is the cheap part, judgment is not. The deck an agent produces in ninety seconds still represents your company when it hits a client's screen, so FlowRunner treats generation as a draft step and review as the gate. - Turn briefs, outlines, and reports into themed decks without design work - Choose text handling per job: generate new copy, condense long input, or preserve it exactly, with tone, audience, and language controls per generation - Export PPTX, PDF, or PNG for the systems and people that need files, not links - Control image sourcing per job, from AI generated to stock sources to no images at all - Track credits deducted and remaining on every completed generation

Without FlowRunner

Every deck starts at slide zero The kickoff deck, the QBR, and the one-pager each begin as a blank template and an evening of formatting

Content trapped in the wrong shape The research doc exists, the proposal exists, but turning either into a presentable deck is its own project

Design debt per presenter Ten people produce ten visual dialects, and brand consistency depends on who had time

With FlowRunner

Decks generated from source content Briefs, outlines, and reports become themed decks through one flow, with copy generated, condensed, or preserved verbatim as the input deserves

Format is a parameter The same content ships as a presentation, a document, a webpage, or a social post by changing one field

Consistency by construction Every generated artifact uses the house theme, tone, and audience settings, because the flow sets them

Use Case Scenarios

Intake form to pitch deck before the kickoff call

A project brief lands via [Typeform](/integrations/typeform). The agent maps the responses into a structured outline, calls Generate & Wait with the house theme, Text Mode set to Generate, and a ten-card limit, and receives the Gamma URL and PPTX export in one blocking step. The account lead reviews the deck, and on approval the link posts to [Slack](/integrations/slack) while the PPTX archives to [Google Drive](/integrations/google-drive). The team walks into kickoff with a branded deck nobody stayed late to build.

The weekly report that becomes a readable document

Every Friday, a flow gathers the week's metrics and status notes into a single text block, then calls Create Generation with Format set to Document and Text Mode set to Condense, so the sprawling input becomes a tight, structured readout. Get Generation polls to completion, and the PDF export goes to stakeholders while the generation's credits deducted and remaining log to [Google Sheets](/integrations/google-sheets). Reporting stops being a formatting chore and becomes a pipeline with a cost line the ops lead can see.

One announcement, four formats

A product update needs a launch deck for sales, a webpage for the changelog audience, and a social post. The agent runs the same source copy through three generations, varying only the Format and Text Amount parameters, with Text Mode set to Preserve where legal reviewed the wording. Each output returns its own shareable URL, and the marketing lead reviews the set together before anything publishes. The announcement ships consistent across every surface because one flow produced every version from one source of truth.

Human-in-Loop Highlight

Create Generation deducts credits from a finite balance every time it completes, and the deduction scales with format, card count, and image generation, so a retry loop with a malformed prompt or an over-eager batch flow can quietly drain the pool your team's real decks depend on. Every completed generation reports credits deducted and remaining, and FlowRunner flows read both. Single generations for a live deliverable run without ceremony, but two things always stop for a person: batch runs, where the ops owner approves the job count and projected credit consumption before the loop starts, and anything that ships outward, because a generated deck carries generated claims, and Text Mode set to Generate means the AI wrote sentences nobody on your team has read yet. The account lead reviews the Gamma link before the URL leaves the building. Generation at machine speed, spend and speech at human discretion.

Agent processes routinely

Detects exception requiring judgment

Clear match Continues automatically

Ambiguous Routes to human via preferred channel

Human decides

Agent resumes with decision

Agent Capabilities

3 actions

Generations

Create Generation Starts an asynchronous Gamma generation from a text prompt or source content. Choose the output format, presentation, document, webpage, or social, how the AI treats your text, generate, condense, or preserve, plus card count, theme, image sourcing, and an optional export format. Returns a generation ID immediately. Requires a paid Gamma plan with API access, currently in beta.
Get Generation Retrieves the status and result of a generation by ID. Poll until status is completed, at which point the response carries the shareable Gamma URL, the export or download URL when an export was requested, and the credits consumed. A failed status includes the error message.
Generate & Wait Creates a generation and polls to completion in one blocking call, with the same inputs as Create Generation plus a timeout. Returns the Gamma URL, any export URL, and credits used. Generation can take a minute or more, so prefer Create Generation plus Get Generation in latency-sensitive flows and reserve this for flows where a single blocking step is acceptable.

Frequently Asked Questions

What can FlowRunner do with Gamma?

FlowRunner agents can run Create Generation, Get Generation, and Generate & Wait in Gamma.

Does connecting Gamma to FlowRunner require OAuth?

No. Gamma connects to FlowRunner with an API key, no OAuth flow required.

Can Gamma trigger a FlowRunner workflow automatically?

Gamma doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.

Start building with Gamma

$100 in credits. No card required. Connect in minutes.

Start Building Free Talk to us