You upload a voice memo, a raw recording of a song’s chorus. The interface is quiet, almost presumptuous. A few clicks later, you’re watching a video: a character walks through a rain-slicked city, the camera pulling back in a slow dolly shot that matches the swell of the music. The scene wasn’t prompted; it was inferred. This is the first impression Koyal wants to make, a 45-second free clip that demonstrates its core proposition: not just generating video, but directing it [Y Combinator Launches, Jan 2025].
A bet on agentic filmmaking
Koyal’s wedge is the shift from prompt engineering to agentic orchestration. Where tools like Runway or Pika ask users to describe a scene, Koyal asks for the raw material,audio, a script, a podcast,and claims its AI handles the rest. The system, according to its launch materials, uses multimodal models to extract emotional tone, then autonomously storyboards shots, shifts camera angles, and maintains character consistency to produce a short cinematic sequence [Y Combinator Launches, Jan 2025]. The target is the creator who has the audio but lacks the time, budget, or technical skill for traditional video production: musicians, podcasters, educators, and startups needing launch reels. The founders, Mehul and Gauri Agarwal, are Carnegie Mellon computer science graduates who built the platform with a team drawn from MIT, Meta, and Carnegie Mellon research [Carnegie Mellon University News]. Their technical foundation includes CHARCHA, a personalization engine presented at the NeurIPS 2024 conference, which the company describes as a secure protocol for tailoring AI-generated content [Business Standard, 2025].
The early traction and partnership play
Despite its recent 2025 founding, Koyal has moved quickly to establish commercial footholds, particularly in media-heavy industries. The company has concluded paid pilot programs with a roster of established names, including Universal Music, T-Series, Maddock Entertainment, and the Collective Artists Network [Y Combinator Entertainment, 2026]. This early focus on labels and production houses, rather than solely individual creators, suggests a strategic push toward the enterprise end of content creation. A separate, announced partnership with Offbeet Media Group at the WAVES Summit 2025 further signals an intent to embed its technology within professional production workflows [Business Standard, 2025]. The table below outlines the key early relationships that form the basis of Koyal’s market entry.
| Partner / Pilot | Type | Context |
|---|---|---|
| Universal Music | Music Label | Paid pilot concluded [Y Combinator Entertainment, 2026] |
| T-Series | Music Label | Paid pilot concluded [Y Combinator Entertainment, 2026] |
| Maddock Entertainment | Production House | Paid pilot concluded [Y Combinator Entertainment, 2026] |
| Collective Artists Network | Talent Agency | Paid pilot concluded [Y Combinator Entertainment, 2026] |
| Offbeet Media Group | Media Group | Strategic partnership announced [Business Standard, 2025] |
Navigating a crowded and capable field
The ambition is clear, but the competitive landscape is dense and evolving rapidly. Koyal operates in a space defined by well-funded incumbents and rapid model advancements. Its differentiation rests on a full-stack, agentic approach to a specific workflow,turning audio into narrative video,rather than being a general-purpose video model.
- The model makers. Competitors like OpenAI’s Sora represent the raw capability frontier. Koyal’s bet is that most users don’t want to become expert prompt cinematographers; they want a director. Its value is the layer of creative automation on top of the underlying models.
- The tool builders. Platforms like Runway and Pika have massive community traction and are iterating rapidly on their own suite of creative controls. Koyal’s narrower focus on a single input-to-output pipeline could be a strength in onboarding specific audiences, but it also limits its surface area.
- The quality ceiling. The ultimate constraint for any AI video tool is the perceptual quality and coherence of the output. While Koyal’s research pedigree is a signal, the product’s success hinges on the output being not just automated, but good,compelling enough for a musician to use in a lyric video or a podcaster to use for a trailer.
The company’s reported $500,000 seed round, raised during its Y Combinator F25 batch, provides runway but is modest relative to the capital deployed in this category [Extruct AI]. Its path will depend on proving that its automated director can achieve a consistency and creative flair that manual prompting cannot, thereby creating a new slot in the creator’s toolkit.
For now, the product lives at its public beta, a quiet invitation to test the premise. You provide the sound; it provides the mise-en-scène. The implicit question Koyal is asking isn’t just about automating filmmaking. It’s about whether the next generation of visual storytelling starts with a voice, a melody, or a line of dialogue,and whether the machine listening can become a worthy collaborator.
Sources
- [Y Combinator Launches, Jan 2025] Koyal: The Agentic AI Filmmaking Platform | https://www.ycombinator.com/launches/ObJ-koyal-the-agentic-ai-filmmaking-platform
- [Carnegie Mellon University News] CMU Alumni Launch Koyal for Safe AI Video Creation | https://www.ri.cmu.edu/cmu-alumni-launch-koyal-for-safe-ai-video-creation/
- [Business Standard, 2025] Koyal announces partnership with Offbeet Media Group, details CHARCHA engine | https://www.business-standard.com
- [Y Combinator Entertainment, 2026] Entertainment Startups funded by Y Combinator | https://www.ycombinator.com/companies/industry/entertainment
- [Extruct AI] Koyal Funding: $1M | Complete Analysis | https://www.extruct.ai/hub/koyal-ai/