Wondercraft.ai

Wondercraft.ai builds AI tools for creating and editing audio and video. Its January 2026 product announcements describe a studio combining third-party generative models, a timeline editor, and Wonda, an agent that edits through natural-language instructions. The company’s earlier podcast builder turned written content into scripts and synthetic speech; its expanded studio retains audio production alongside video.[1][5][6]

The video studio followed two earlier product generations: a podcast builder launched in 2023 and a broader audio studio launched in January 2024. Those products established the script, voice, music, and publishing workflows that preceded the video expansion.[4][1][5]

Youssef Rizk and Dimitris Nikolaou met while working as engineers at Palantir. Wondercraft’s Y Combinator profile lists a 2023 founding date and the Summer 2022 batch, alongside their previous venture, Moonshot, which offered investment in athletes. In his September 2023 interview, Rizk described Moonshot as a legal and regulatory product. He recalled concluding that it lacked demand and did not fit the founders’ strengths, prompting a pivot toward technical work.[2][1]

The founders tested the podcast idea with a one-page React prototype without authentication. After a Twitter post brought inbound interest, they added a hurried Stripe payment flow. In his September 2023 interview, Rizk recalled making approximately “three K” in one day while charging “50 bucks a month.” This was an early willingness-to-pay test; his account did not specify the currency or establish sustained revenue.[1]

Oskar Serrander, formerly COO of Acast and an executive at Spotify and iHeartMedia, subsequently joined forces with the engineers as a cofounder. In his February 2024 account, Serrander recalled receiving a cold direct message from Nikolaou after leaving Acast, seeing the existing MVP, and developing the company’s ambitions together. He dated the first product launch to May 2023.[3][4]

By September 2023, Wondercraft combined optional script generation with AI voice production, using GPT-3/4 and ElevenLabs. Users could convert blogs or newsletters into audio, specify pronunciations, set recurring introductions, outros, music, and speakers, and publish through hosted RSS feeds. A newly announced Zapier integration could pick up newsletter posts from an RSS feed and generate episodes automatically. Users could review and modify the generated episode before publishing, or schedule its release.[1]

Hacker News Recap demonstrated the workflow as an automated daily digest of the top ten Hacker News posts. The company’s historical YC launch reported reaching Spotify’s top 40 technology podcasts within one month; Rizk reported a top-30 technology position in the September 2023 interview. Wondercraft also produced audio versions of Paul Graham’s essays. By that interview, it had dropped its Product Hunt Daily show.[2][1]

Rizk argued that Wondercraft’s value lay in a complete podcast workflow: editing, music, publication, and reusable settings around general-purpose speech APIs. His voice-model evaluation emphasized realism and controllability, including predictable segment duration and reliable tone. In the September 2023 interview, he summarized his restraint on GPU spending as “GPUs after PMF,” explaining that such investment needed to justify its effect on runway.[1]

Wondercraft introduced translation and dubbing in 2023. For video, it aligned translated audio with speaker turns, keeping the dubbed speaker matched to the person on camera rather than synchronizing lip movements. Rizk said professional translators provided quality assurance, addressing the difficulty of judging a dub in a language the creator does not speak. January 2024 reporting described dubbing into more than 15 languages, supported by over 100 language experts who edited, validated, and checked translations.[1][3]

Version two launched on January 29, 2024, broadening Wondercraft into an audio studio for podcasts, audiobooks, advertisements, meditations, narration, and company communications. It combined templates, AI script assistance, music, synthetic voices, video generation, and translation. Serrander described both a free offering and monthly subscriptions, with larger enterprise customers beginning onboarding in 2024.[4][3]

On January 29, 2024, Wondercraft announced that it had raised a $3 million seed round, led by Will Ventures and joined by ElevenLabs and Steven Bartlett, following initial Y Combinator funding. Tech.eu’s report also put usage at more than 30,000 creators, writers, entrepreneurs, advertisers, and brands after the 2023 beta launch. That aggregate did not specify paying customers or active users.[3]

CoinDesk Podcasts supplied a concrete production example. Its executive producer, Jared Schwartz, said Wondercraft had been integrated into its shows “for a more efficient process without sacrificing quality.” The testimonial described workflow adoption without quantifying the efficiency improvement.[3]

In an announcement dated January 27, 2026, Nikolaou described a decision made about a year earlier to build a video studio for teams. He argued that generating clips had become easier while turning them into editable, repeatable, branded productions remained difficult. Wondercraft consequently organized creation around tasks such as training, onboarding, product launches, social advertisements, and refining existing footage, primarily targeting marketing, communications, enablement, HR, and learning and development teams.[6]

The expanded studio has three company-described layers: models, an editor, and workflows powered by Wonda. Named integrations include Runway, Luma, VEO3, and ElevenLabs. Wonda orchestrates creation and accepts instructions to cut, trim, animate, add captions, and incorporate music or voice; the timeline editor provides control over timing and branding. Users can begin with footage, a document, a script, or an idea, generate a first draft, and refine it. Audio remains part of the studio, including turning podcasts or ad scripts into video.[5][6]