Koyal

Agentic AI platform turning audio/scripts into cinematic videos

Website: https://koyal.ai

Cover Block

PUBLIC
Name Koyal
Tagline Agentic AI platform turning audio/scripts into cinematic videos
Headquarters San Francisco, CA, USA
Founded 2025
Stage Seed
Business Model SaaS
Industry Media / Entertainment
Technology AI / Machine Learning
Geography North America
Growth Profile Venture Scale
Founding Team Co-Founders (2)
Funding Label Seed (total disclosed ~$500,000)

Links

PUBLIC

Executive Summary

PUBLIC Koyal is an early-stage platform attempting to automate cinematic video production by converting audio or scripts into fully realized short films, a process that currently requires significant manual labor and specialized skills [Y Combinator Launches, Jan 2025]. The company's proposition centers on an "agentic" workflow that handles storyboarding, character consistency, and camera direction, positioning it as a potential tool for content creators and small production houses rather than a raw model competitor [Carnegie Mellon University News, post-2025].

Founded in 2025 by Carnegie Mellon School of Computer Science graduates Mehul and Gauri Agarwal, the company leverages academic research, including a patented personalization protocol presented at NeurIPS 2024 [Business Standard, 2025]. The team's composition, drawing from institutions like MIT and Meta, suggests a research-heavy foundation, though its commercial execution remains unproven [Y Combinator Entertainment, 2026].

Koyal participated in Y Combinator's F25 batch and has disclosed a seed round of approximately $500,000, though the lead investor and valuation are not public [Extruct AI, post-2025]. Its business model is SaaS, with initial go-to-market focused on a free beta offering to attract musicians, podcasters, and filmmakers [LinkedIn, 2026].

The critical watchpoints over the next 12-18 months are the conversion of announced pilot programs with entities like Universal Music into recurring revenue, the technical delivery of its promised agentic pipeline against established video generation tools, and its ability to articulate a clear monetization strategy beyond a free trial.

Data Accuracy: YELLOW -- Core company claims and team background are confirmed by multiple sources, but funding details and commercial traction lack independent corroboration.

Taxonomy Snapshot

Axis Value
Stage Seed
Business Model SaaS
Industry / Vertical Media / Entertainment
Technology Type AI / Machine Learning
Geography North America
Growth Profile Venture Scale
Founding Team Co-Founders (2)
Funding Seed (total disclosed ~$500,000)

Company Overview

PUBLIC

Koyal is a venture-scale AI video generation startup founded in 2025 by Carnegie Mellon University computer science graduates Mehul and Gauri Agarwal. The company, which is headquartered in San Francisco, was launched publicly as part of the Y Combinator F25 batch in January 2025 [Y Combinator Launches, Jan 2025]. Its core proposition is an agentic platform that automates the conversion of audio inputs into cinematic video, a process the founders developed from academic research presented at the NeurIPS conference [Business Standard, 2025].

The company's early trajectory is marked by a series of structured commercial and technical milestones. Following its YC launch, Koyal initiated paid pilot programs with several major media and music entities, including Universal Music, T-Series, Maddock Entertainment, and the Collective Artists Network [Y Combinator Entertainment, 2026]. It also announced a strategic partnership with Offbeet Media Group at the WAVES Summit in 2025 [Business Standard, 2025]. On the product side, the team has publicly released a beta version of its platform at beta.koyal.ai [Y Combinator Entertainment, 2026].

Data Accuracy: YELLOW -- Key founding and accelerator details are confirmed, but some commercial milestones are sourced from a single company-associated page.

Product and Technology

MIXED

Koyal's platform is built around a single, ambitious workflow: converting an audio file or script into a fully produced, cinematic video. The process is designed to be agentic, meaning the AI handles tasks traditionally requiring human direction, such as storyboarding, editing, and managing camera angles [Y Combinator Launches, Jan 2025]. The system uses multimodal AI models to extract emotional tone from the input audio, which then informs the visual style, character expressions, and scene transitions [Carnegie Mellon University News, post-2025]. This approach aims to generate consistent characters and personalized scenery across a video sequence, a technical challenge for many AI video tools.

The core technology differentiator appears to be the CHARCHA personalization engine, which the company describes as a "patented secure personalization protocol" and a "genAI captcha" [Y Combinator Entertainment, 2026]. This engine was presented at the NeurIPS 2024 conference, suggesting an academic research foundation [Business Standard, 2025]. Publicly, the product offers a free trial for generating 45-second video clips, allowing users to test features like custom avatar creation and dialogue tone editing without a financial commitment [Carnegie Mellon University News, post-2025]. The platform is currently in public beta, accessible at beta.koyal.ai [Y Combinator Entertainment, 2026].

From a technical staffing perspective, the team's composition is a key asset. Development is led by researchers with backgrounds from Carnegie Mellon University, MIT, and Meta, with the co-founders' own academic work at CMU directly informing the product [Carnegie Mellon University News, post-2025] [Y Combinator Entertainment, 2026]. This research-heavy pedigree is intended to translate into more controlled and sophisticated video generation compared to using off-the-shelf foundation models alone. The technology stack is not publicly detailed, but the focus on AI video generation, personalization protocols, and a web-based beta platform suggests a cloud-native architecture built on contemporary machine learning frameworks.

Data Accuracy: YELLOW -- Product claims are sourced from company launch materials and university coverage; technical details like the CHARCHA engine are noted in press but lack independent technical verification.

Market Research

MIXED, The market for AI-generated video is moving from a novelty for hobbyists to a production tool for professionals, driven by a convergence of demand for scalable content and falling costs for computational power.

The total addressable market for AI video generation is not yet formally sized by major research firms. Analysts typically point to the broader creative software and video production markets as proxies. The global video editing software market was valued at approximately $2.5 billion in 2023 and is projected to grow at a compound annual rate of 6.5% through 2030, according to Grand View Research [Grand View Research, 2024]. The adjacent market for digital video content creation, encompassing tools for creators, is significantly larger and growing faster, driven by the expansion of social video platforms and direct-to-consumer media.

Demand is fueled by several tailwinds. The need for video content is exploding across social media, marketing, and entertainment, but traditional production remains costly and time-intensive. This creates a persistent supply gap. Concurrently, advancements in generative AI models for imagery, video, and audio are rapidly improving output quality and coherence, lowering the technical barrier to entry. A third driver is the professionalization of the creator economy, where individual creators and small studios seek broadcast-quality output without the budget for large crews [Y Combinator, 2025].

Key adjacent markets include traditional video production services, stock footage libraries, and animation studios. AI video platforms like Koyal position themselves as substitutes for low-to-mid-fidelity segments of these markets, automating tasks like storyboarding, basic animation, and scene generation. The regulatory landscape is nascent but evolving, with increasing focus on copyright, deepfake disclosure, and data privacy, which could impact training data sourcing and output usage [Business Standard, 2025].

Metric Value
Video Editing Software (2023) 2.5 $B
Projected CAGR (to 2030) 6.5 %

While the direct TAM for agentic AI filmmaking is unquantified, the growth trajectory of the underlying video creation ecosystem suggests a substantial runway. The primary constraint is not demand, but the technology's ability to reliably meet professional quality and consistency requirements at scale.

Data Accuracy: YELLOW, Market sizing is drawn from an analogous, broader sector report. Demand drivers are inferred from industry coverage and company positioning.

Competitive Landscape

MIXED Koyal enters a crowded field of AI video generators by positioning itself not as a raw model but as an automated production pipeline for narrative content. The competitive map can be segmented by technical approach and target user.

Direct competitors in generative AI video are well-funded and moving fast. Runway has established a strong foothold with creative professionals through its suite of editing tools and its Gen-2 model, raising over $190 million [Crunchbase]. Pika Labs, known for its accessible text-to-video interface, has also secured significant venture backing. OpenAI's Sora, while not yet publicly available, represents the frontier of model capability from a major platform player. These companies compete on the quality and controllability of the raw video output itself.

Adjacent substitutes include traditional video editing software like Adobe Premiere Pro, which integrates AI features like Firefly, and a host of specialized tools for scriptwriting, storyboarding, and audio production. For Koyal's target users, these represent the incumbent workflow its platform aims to consolidate and automate.

Company Positioning Stage / Funding Notable Differentiator Source
Koyal Agentic pipeline from audio/script to cinematic short films Seed (~$500k) Focus on narrative consistency, automated storyboarding, and emotion extraction from audio [Y Combinator, Jan 2025]
Runway AI-powered creative suite for video editing and generation Series C Established toolchain, strong brand with professional creators, multi-modal editing features [Crunchbase]
Pika Text-to-video generation platform emphasizing ease of use Venture User-friendly interface, rapid iteration on community-driven features [Crunchbase]
Sora (OpenAI) High-fidelity video generation from text prompts Corporate R&D Exceptional output quality and physics simulation from a leading AI research lab [OpenAI]

Koyal's stated defensible edge today is its research-backed focus on the full narrative pipeline, specifically its CHARCHA personalization engine and its emphasis on extracting emotional tone from audio to guide visual generation [Business Standard, 2025] [Y Combinator, Jan 2025]. This edge is rooted in the team's academic background from Carnegie Mellon and MIT, where foundational research was presented at NeurIPS [Business Standard, 2025]. However, this edge is perishable. The core AI models for video generation are rapidly commoditizing, and larger competitors can easily replicate a pipeline wrapper if the underlying narrative logic proves valuable. Koyal's durability will depend on the depth of its proprietary datasets for character consistency and the sophistication of its directorial agents, which are not yet publicly demonstrated at scale.

The company is most exposed on two fronts. First, it lacks the raw model horsepower and compute resources of players like OpenAI or Runway, which could allow them to leapfrog on output quality. Second, its go-to-market targets a niche of creators needing narrative shorts, a segment that may be too small to support a standalone venture if general-purpose tools add similar storytelling features. Its distribution is nascent, relying on a public beta and Y Combinator's network, while competitors own established creator communities and app store placements.

The most plausible 18-month scenario is one of rapid feature convergence. If general video models from Runway or Pika integrate basic script-to-video templating and character consistency controls, Koyal's unique value proposition narrows significantly. The winner in this scenario would be the platform that best balances high-quality generation with an accessible, integrated workflow for storytellers, which could still be an incumbent. Koyal becomes a loser if it cannot transition from a promising research project to a product with clear, measurable superiority in either output quality or user workflow efficiency before the larger players close the feature gap. Its survival likely hinges on securing a partnership with a major content studio or music label to create a vertical-specific, defensible use case, a path its concluded pilots with Universal Music and others may be testing [Y Combinator Entertainment, 2026].

Data Accuracy: YELLOW -- Competitor funding and positioning are from public databases; Koyal's differentiators are cited from launch materials but lack third-party validation of technical superiority.

Opportunity

PUBLIC

If Koyal's technology can reliably automate the most labor-intensive parts of video pre-production, the prize is a share of the global content creation economy, valued in the hundreds of billions, currently bottlenecked by cost and expertise.

The headline opportunity is Koyal becoming the default storyboarding and pre-visualization layer for professional media production. The outcome is reachable because the company is not positioning itself as a general-purpose video generator, but as an "agentic" platform that handles the specific, structured workflow of turning a script or audio into a cinematic storyboard with consistent characters and shifting camera angles [Y Combinator Launches, Jan 2025]. This addresses a real pain point: directors and producers spend significant time and money on manual storyboarding and animatics to secure funding and align creative teams. Koyal's early engagement with major music labels and production houses, including concluded pilots with Universal Music and Maddock Entertainment [Y Combinator Entertainment, 2026], demonstrates that its value proposition is being tested at the professional tier, not just by hobbyists. Success here would mean Koyal's output becomes a non-negotiable step in the pitch and planning process for films, music videos, and serialized content.

Growth from a promising tool to a scaled platform could follow several concrete paths.

Scenario What happens Catalyst Why it's plausible
API-as-a-Service for Music Koyal's audio-to-video engine is embedded into the internal tools of major record labels to rapidly produce visualizers and lyric videos for new releases. A formal, scaled partnership with a top-three global music label following the pilot with Universal Music. The company has already concluded a pilot with Universal Music Group [Y Combinator Entertainment, 2026], establishing a beachhead with a label that has a vast, continuous release schedule needing cost-effective visual content.
The "Canva for Video" for Creators The public beta platform sees viral adoption among indie musicians, podcasters, and educators, who use it to produce promotional and educational content, driving a freemium-to-paid conversion funnel. A key feature or pricing tier launch that dramatically lowers the barrier for high-quality output, coupled with influencer-led marketing. The product's initial wedge is a free 45-second video clip for new users [Carnegie Mellon University News], a classic top-of-funnel user acquisition strategy for creator tools. The target buyer profile explicitly includes musicians and podcasters [Y Combinator Launches, Jan 2025].
Standardization in Indian Media Koyal becomes the go-to AI pre-vis tool for India's massive film and television industry (Bollywood, regional cinema, streaming originals), leveraging local founder connections and early pilot partners. A strategic partnership with a major Indian studio or streaming platform, such as the announced partnership with Offbeet Media Group [Business Standard, 2025], expanding to multiple projects. The founders' Indian background and the company's cited pilot with T-Series and Collective Artists Network [Y Combinator Entertainment, 2026] provide a direct entry point into one of the world's most prolific content ecosystems.

Compounding for Koyal would manifest as a data and workflow moat. Each project completed on the platform generates proprietary data on how narrative beats, emotional tone, and directorial intent ("shift camera angles here") map to specific visual outputs. This dataset, focused on cinematic grammar rather than generic imagery, would be difficult for competitors using off-the-shelf models to replicate. The company's research into a "patented secure personalization protocol CHARCHA" [Y Combinator Entertainment, 2026], presented at NeurIPS 2024 [Business Standard, 2025], suggests an early focus on building defensible, model-level differentiation from user data. Furthermore, as professionals integrate Koyal's storyboards into their standard workflow, switching costs rise; the platform becomes the shared visual language for a production, embedding itself into the creative process.

The size of the win can be framed by looking at comparable companies that have automated a layer of creative production. For instance, Canva, which simplified graphic design, reached a peak valuation of approximately $40 billion. A more direct, though earlier-stage, comparison is Runway, an AI video generation toolset, which was valued at $1.5 billion in its 2023 Series C. If Koyal successfully executes the "API-as-a-Service for Music" scenario and captures a meaningful portion of the professional pre-visualization market, a valuation in the low single-digit billions is a plausible outcome (scenario, not a forecast). This is supported by the significant capital flowing into generative AI media tools and the high willingness of media companies to pay for solutions that reduce production time and cost.

Data Accuracy: YELLOW -- Opportunity scenarios are extrapolated from cited pilot programs and stated target markets; specific commercial terms and scale of early engagements are not publicly quantified.

Sources

PUBLIC

  1. [Y Combinator Launches, Jan 2025] Koyal: The Agentic AI Filmmaking Platform | https://www.ycombinator.com/launches/ObJ-koyal-the-agentic-ai-filmmaking-platform

  2. [Carnegie Mellon University News, post-2025] CMU Alumni Launch Koyal for Safe AI Video Creation | https://www.ri.cmu.edu/cmu-alumni-launch-koyal-for-safe-ai-video-creation/

  3. [Business Standard, 2025] Strategic partnership with Offbeet Media Group at WAVES Summit 2025 | https://www.business-standard.com/ (URL from structured facts)

  4. [Extruct AI, post-2025] Koyal Funding: $1M | Complete Analysis | https://www.extruct.ai/hub/koyal-ai/

  5. [Y Combinator Entertainment, 2026] Entertainment Startups funded by Y Combinator (YC) 2026 | https://www.ycombinator.com/companies/industry/entertainment

  6. [LinkedIn, 2026] Kush Taneja - Servant to the young | https://www.linkedin.com/in/iamkushtaneja/

  7. [Grand View Research, 2024] Video Editing Software Market Size Report, 2024-2030 | https://www.grandviewresearch.com/ (URL inferred from standard report naming)

  8. [Crunchbase] Runway Company Profile | https://www.crunchbase.com/

  9. [OpenAI] Sora: Creating video from text | https://openai.com/sora

  10. [Y Combinator, Jan 2025] This YC Startup Will Take Over Hollywood: Koyal (YC F25) | https://www.youtube.com/watch?v=GbWgZdRN7PM

Articles about Koyal

View on Startuply.vc