Gamma

Gamma's core loop — paste an outline, a PDF, or a prompt and get a structured, on-theme presentation, document, website, or social post back in under a minute — is the product's real strength, and it's the reason it has grown past 100 million users by the vendor's own count (gamma.app/about, self-reported, not independently audited). It ships one API (POST /generations) that covers all four output formats plus PDF/PPTX/PNG export and Google Slides export, with an MCP server and native Zapier/Make/n8n integrations already live (developers.gamma.app/reference/changelog.md), so teams that want to generate decks or docs from another tool's data don't have to build a custom exporter. It draws on a genuinely broad model roster — Claude, GPT/DALL-E, Gemini/Imagen, Flux, Recraft, Ideogram, Luma, Leonardo — rather than being locked to one vendor's output style (gamma.app/pricing).

Pricing is public and not gated behind a 'contact sales' wall. Free gives 400 one-time signup credits and 10-slide generations. Plus is $9/seat/month ($108/year) with 1,000 credits/month; Pro is $18/seat/month ($216/year) with 4,000 credits/month, custom branding, and API access; Ultra is $90/seat/month ($1,080/year) with 20,000 credits/month and the most advanced models. Team and Business plans (not shown with public prices) get 6,000 and 10,000 credits/seat/month respectively (gamma.app/pricing, verified 2026-08-27). Unused credits roll over up to 2x the monthly allotment. The trade-off buyers should weigh: this is consumption pricing dressed as a seat price — how far a plan actually goes depends on which AI model a generation happens to use, and several 2026 Trustpilot reviewers report the system silently defaulting to pricier models and burning through credits faster than expected (trustpilot.com/review/gamma.app, review dated 2026-03-31).

Gamma raised a $68 million Series B in November 2025 at a $2.1 billion valuation, led by Andreessen Horowitz with Accel and Uncork Capital participating; the round included a $20 million secondary for employee liquidity. The company says it was profitable and had reached $100 million ARR and 70 million users at announcement, achieved with roughly 50 employees (TechCrunch, Julie Bort, 2025-11-10, citing founder/CEO Grant Lee's public statements — these growth and profitability figures are the company's own claims, not independently audited). Total funding is roughly $90 million since a $7 million seed (2021) and a $12 million Accel-led Series A (2024). As of August 2026, Gamma acquired Accel-backed startup Lica to build an internal 'design research' team focused on more interactive, multimodal presentation formats (TechCrunch, Ivan Mehta, 2026-08-25) — a signal the product is expanding beyond static slides, and also that some near-term engineering effort will go to R&D rather than fixing the editing complaints below.

Gamma is rated 'Bad,' 1.7 out of 5, on Trustpilot as of 2026-08-27, based on reviews spanning June 2025 through August 2026 (trustpilot.com/review/gamma.app). The recurring complaints are not about the initial generation — several reviewers say the first draft looks fine — but about the AI editing agent failing to make targeted changes ('I asked them to change one slide, and they changed all'), inability to control layout precisely enough for print/PDF formats, and a cluster of billing disputes describing being charged for annual renewals they say they didn't intend to keep and then being refused refunds. One reviewer (2026-05-12) pushed back on the pattern, saying the negative reviews reflect users who didn't invest in prompt engineering rather than a broken product. This is a real tension buyers should weigh against the vendor's growth numbers above: a tool can be growing fast on the strength of its generation step while frustrating a meaningful share of paying users once they try to refine the output.

Gamma publishes a SOC 2 Type II attestation through a Vanta-hosted trust center (trust.gamma.app) and a public Data Processing Addendum (gamma.app/dpa). Its subprocessor list is unusually long and worth reading before sending sensitive content through it: 35 named subprocessors as of 2026-08-27, including nine separate AI-model providers (Anthropic, OpenAI, Google, Microsoft, Black Forest Labs, Ideogram, Leonardo, Luma, Recraft, Perplexity) plus infrastructure, support, and analytics vendors (gamma.app/subprocessors). That breadth is a direct consequence of the multi-model design that makes the product good — but it also means content pasted into Gamma can transit more third parties than a single-model competitor.