Synthesia's AI-avatar bet: enterprise video from text, $100M ARR, $4B valuation

London-based B2B AI video platform: turn text documents into presenter-led videos with AI avatars, used for corporate training, sales and marketing; 240+ avatars across 160+ languages.

That AI would push video creation cost to zero and avatars in 160+ languages would make video the default way businesses train, sell and communicate — replacing studios.

$180M Series D at a $2.1B valuation (Jan 2025, NEA-led); $200M Series E at $4B (Jan 2026, GV-led); $330M raised in total as of Jan 2025 (TechCrunch)

Founded in London in 2017 by Victor Riparbelli, Steffen Tjerrild, TU Munich professor Matthias Niessner and UCL professor Lourdes Agapito. Riparbelli's core beliefs: AI would drive the cost of content creation to zero, and video was a better way for organizations to communicate — if only production were not so expensive. The team built avatar video generated from text and aimed it squarely at business users.

By June 2023 Synthesia had raised a $90M round at a $1B valuation with Kleiner Perkins and Accel and grown to 50,000 business customers. In January 2025 NEA led a $180M Series D at a $2.1B valuation — with WiL, Atlassian Ventures, PSP Growth and GV — on 60,000 customers and 1 million users; by April 2025 the company passed $100M ARR and took investment from Adobe as it began building AI agents for learning. In January 2026 Google Ventures led a $200M Series E at a $4B valuation, with an employee secondary sale arranged with Nasdaq.

Still scaling: as of the January 2026 Series E, Synthesia is one of the UK's most valuable private AI companies, says over 90% of the Fortune 100 uses its platform, and is pushing into conversational AI agents for enterprise learning after passing $100M ARR.

Enterprises already paid studios and agencies for training video, so a good-enough AI avatar directly replaced a large recurring cost — a wedge consumers would not pay for.

Focusing on business use kept Synthesia out of the deepfake controversy that shadowed consumer avatar apps and let it win safety- and compliance-conscious buyers.

The numbers compounded: 50,000 customers (2023) to 60,000 and 1M users (2025), $100M ARR (2025), then a $4B round (2026) — evidence that enterprise AI can scale revenue before profitability.

The 2026 pivot to AI agents kept the thesis alive: the same content-cost argument now applies to interactive learning, not just rendered video.

Sell the boring use case first: good-enough avatars for training beat flashy deepfakes. A cost-to-zero bet pays where customers already pay heavily for video — then expand from that wedge.

As of January 2026 Synthesia remains independent and scaling from London, with offices across Europe and the US. It raised $200M at a $4B valuation — double the $2.1B of a year earlier — with Google Ventures leading and NVentures, Accel, Kleiner Perkins, NEA, PSP Growth, Air Street, FirstMark and MMC participating, plus an employee tender at the same valuation. Its next bet is conversational AI agents for enterprise learning and upskilling, and the company says over 90% of the Fortune 100 already uses its platform.