each::labs, a San Francisco generative media platform founded by engineers from Türkiye, has raised a $7 million seed round co-led by Revo Capital and the European Bank for Reconstruction and Development (EBRD). The company announced the round on 18 November 2026 and says its annualised revenue run rate passed $20 million in October on less than $2 million raised before.
Quick Take
- Round: $7 million seed, co-led by Revo Capital and the EBRD.
- Traction: monthly revenue grew about 5x between August and October to roughly $2 million; more than 750 organisations paid in October.
- Scale: about one million generation runs a day in October, with a team of about 20.
- Product: images, video and audio from its own models plus more than 600 third-party models, via API, the each::studio workspace or plugins for AI agents.
- Use of funds: more in-house models, sales and marketing hires in San Francisco and a bigger each::studio.
In plain English
Making AI images or videos often takes many attempts, and each try costs money. each::labs sells a single platform that connects to hundreds of AI image, video and sound models and handles the messy parts, like retrying when a model fails. Businesses such as apps, online shops and ad companies use it to produce media at scale. It also builds its own models when the available ones are too expensive.
What happened
each::labs announced the seed round on its blog, and The Next Web and Tech.eu reported it the same day. Co-founder and CEO Eftal Yurtseven said: “Before generative media, if you couldn’t draw, you couldn’t show anyone what was in your head. Now you can, but only after fighting the tool for it.” Revo Capital founding partner Cenk Bayrakdar said: “We watched each::labs’ customers go from running tests to running their products on it. That is the moment a platform becomes infrastructure.”
Background
Yurtseven and co-founder Canberk Sinangil founded the company in June 2024; both previously worked as engineers at Insider. Right Side Capital led a first round in February 2025, and Revo Capital first invested in September 2025. In July the company rebuilt its platform: on replayed customer traffic, the success rate rose from 56% to 100% and core requests ran up to 25 times faster, according to each::labs.
How it works
- One API for many models: customers route image, video and audio jobs across more than 600 models with fallbacks when a generation fails.
- Workflows: teams build a pipeline once and run it at volume, with spending alerts.
- Own models: Flint Image returns four images for the price of one; Lockstep, a recasting model that keeps a clip’s motion and swaps the cast, launched on 28 October at $0.09 per second of 720p video.
- Customers: ad platform Wask uses it for AI creatives and produces more than 1,000 a minute at peak.
Why it matters
For developers building AI media apps, aggregation layers like each::labs reduce the work of keeping up with new video and image models every week.
For the generative media market, the numbers show how quickly consumer AI apps moved into video this autumn, and how much demand sits in tasks such as character swapping.
For Europe and Türkiye’s startup scene, the EBRD co-leading a seed round in an AI media company is a notable signal.
What’s still open
- Margins on reselling third-party model access as model prices fall.
- How each::labs competes with larger inference platforms that also host image and video models.
- Rights and consent checks for recasting tools like Lockstep.
Sources
- each::labs, “each::labs raises $7M seed after revenue grows 5x,” company blog, 18 November 2026
- The Next Web, “Generative media startup each::labs raises $7M seed co-led by the EBRD,” 18 November 2026
- Tech.eu, “each::labs bags $7M to make generative media production faster and cheaper,” 18 November 2026
i10x Perspective
The money in generative video is moving to the layer that makes many models usable in production. each::labs is small but growing quickly. Track more rounds in the AI Funding Index and the Funding & M&A section, or browse the latest tools in Video & Image.



