Caroline Clark Launches Arcade 3.0, an AI Video Playground for Marketing Teams
Frontier AI has solved math. Science. Coding. But it still hasn’t solved taste. Until now. Introducing Arcade 3.0: the AI video playground for marketing teams. Type one line, answer two questions, and Arcade does the rest. Voice, captions, and music included. Arcade runs its own harness on top of the best models. It’s seen your actual product AND knows your brand. The numbers show it: in a blind test, B2B marketers chose Arcade’s video 2x as often as videos from Claude Opus 5.5 and Gemini Argon. And you can change one thing without redoing the whole video. Oh, and this video? Made with Arcade. Comment "ARCADE" in the next 48 hours for 50% off your first 2 months of Arcade Pro.
Context
A marketing video tool that claims a blind test win over two frontier models is trying to turn 'taste' into a product, but those test results are the company's own.
Caroline Clark posted that Arcade 3.0 is an AI video playground for marketing teams, where a person types one line, answers two questions, and Arcade does the rest, including voice, captions, and music. She said Arcade runs its own harness on top of the best models, has seen the actual product and knows the brand, and that in a blind test, B2B marketers chose Arcade's video twice as often as videos from Claude Opus 5.5 and Gemini Argon. The post also includes a 48 hour discount prompt. The test result is the company's own claim.
Deep dive
Caroline Clark, posting as the product lead, introduced Arcade 3.0 as an AI video playground for marketing teams. According to the post, a user types one line, answers two questions, and Arcade does the rest, with voice, captions, and music included. She said Arcade runs its own harness on top of the best models, that it has seen the actual product and knows the brand, and that a person can change one thing without redoing the whole video. She also said the attached video was made with Arcade, and offered 50% off the first two months of Arcade Pro to people who comment a keyword in the next 48 hours.
The sharpest claim is a blind test in which, she said, B2B marketers chose Arcade's video twice as often as videos from Claude Opus 5.5 and Gemini Argon. That comparison is Arcade's own, and the post does not describe the sample size, the prompts, or who ran the test. It should not be treated as an independent model leaderboard result.
This is a first hand product launch from the person promoting it, not a lab announcement. The 'taste' framing is marketing language. The useful product idea is a harness that already knows a company's brand and can edit one part of a video without a full rerun.
The existence of Arcade 3.0 and the listed features are confirmed as Clark's announcement. The 2x blind test is an attributed company claim.
What to expect next
A product page or later post should show pricing after the 48 hour offer, who can use Arcade, and a fuller writeup of the blind test. Watch whether marketers can actually edit one element without a full regenerate, and whether the Claude and Gemini comparison is repeated by anyone else. If an official Arcade account or a lab partner posts, that can be added. The test figures should stay labeled as the company's own until then.
