Loading page…
AI video tools for creators and enterprises, monetized through subscriptions, credits, and API usage.
Akool sat near $100K a year in 2021 and 2022 while chasing slow enterprise contracts; the fix was to "just put a payment gateway on it" so self-serve users could pay, and by 2023 revenue was $2–3 million, with Coca-Cola arriving inbound after a three-month benchmark.
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
Jeff Lu left Google Cloud in 2022 to build Akool around generative image and video models. Its first popular product was a family of video edits, including face, character, motion, head, and hair swaps. The company later added avatars, with static and live interaction formats proving popular in B2B, and recently introduced an agent-assisted long-form video canvas. Akool’s strategy spans self-serve prosumers and larger business accounts. Lu said early enterprise contracting moved slowly, so the team opened online payments and gained users who arrived organically. Early acquisition included product demos in Reddit, Facebook, and Discord communities, plus conversations with Fiverr video creators. Coca-Cola became a major enterprise proof point: its team approached Akool without revealing who it was, benchmarked tools for about three months, then ran a six-month face-swap campaign. Reported recognized revenue rose from about $2–3 million in 2023 to $12–13 million in 2024 and $25 million in 2025. Lu said June 2026 revenue exceeded $3 million, with a 2026 revenue expectation of about $60 million. The company has raised about $10 million in total, with its last priced round at roughly a $100 million valuation.
For Akool, the strongest early customer wedge was visible output: face swaps and other edits are easy to demonstrate in a short clip, and the result itself explains the product faster than a feature list.
The quick-reference card — acquisition channels, replication playbook, and risk map — is members-only.
Join the Unicorn Club — $5/moCancel anytime, in one clickPaste into Claude Code / Codex and get a working version of this product end-to-end.
Two production-grade prompts per idea — a full build spec and an SEO growth plan — plus 10 tailored build specs a week. Members only.
Join the Unicorn Club — $5/moCancel anytime, in one clickEvery claim traces back to a listed source — open any of them and check the number yourself.
The full captured transcripts behind these sources are in the Unicorn Club — search them, quote them, check every figure. Join the Unicorn Club
Data credibility: Founder interview with CEO Jeff Lu on Nathan Latka's channel; revenue, customer counts, contract size and usage figures are founder-reported and unaudited. The $36M in the video title is not stated anywhere in the interview; this entry uses the founder's own figures ($25M recognized revenue in 2025, over $3M in June 2026). The interviewer cited Ahrefs traffic data, which is not used here as revenue evidence.
Akool’s first repeatable dollar could arrive through a straightforward online payment, while a large contract required a much longer buying process. The useful lesson for a video generation product is to support both motions from the same product core.
Akool’s reinforcing loop joins model quality, public demonstrations, self-serve usage, and larger deployments. Better task-specific results give creators a reason to show the output; those examples can attract more users from communities already interested in video tools.
For a model-heavy video product, silence can mean weak distribution, slow evaluation, or an output that misses the customer’s standard. Separate those cases before increasing spend.
The rest of this breakdown — 9 more sections — is in the Unicorn Club, with every section of all 1,292 ideas: first customers, first dollar, the flywheel, and the silent stretch.
Join the Unicorn Club — $5/moCancel anytime, in one click