Panjaya

Market position

Panjaya competes in AI-based video translation and dubbing against larger platforms such as Vimeo and ElevenLabs and smaller specialists including Speechify and Synthesis, with third-party profiles also listing Deepdub, HeyGen, Synthesia, Camb AI, Dubverse, Vozo AI and SpeechLab as comparable companies. Coverage notes headwinds for dubbing generally, as captions and subtitles have become a default consumption mode (a CBS poll cited more than half of American TV viewers keeping subtitles on some or all of the time, and TikTok enabling captions by default from November 2023), while also citing research that native-language content drives better engagement, particularly in B2B contexts.

Panjaya's stated differentiator is combining translation and voice cloning with visual adaptation: rather than only replacing audio, its engine modifies lip movements and body language and aligns dubbed audio timing with the original speaker's pauses, breaths and gestures. The CEO says the company uses "best of breed" third-party large language models, translation and text-to-speech services where adequate solutions exist, while its lip-sync engine is built in-house because available tools did not meet its quality requirements for multiple speakers, camera angles and business use cases. Management also positions the product away from the "deepfake" label, describing it instead as "deep real" and citing B2B-only distribution, controlled access and future watermarking as safeguards.

Technology

BodyTalk is a multi-modal pipeline: proprietary algorithms analyze the source video, outputs from third-party translation and text-to-speech services are adapted, and home-grown AI models perform lip syncing. Translation is offered in 29 languages, the dubbed voice mimics the original speaker, and the video is altered so lips and other movements fit the new words and phrasing. The lip-sync engine is developed entirely in-house and is being extended to handle dynamic camera angles, speaker occlusions and interactions between multiple speakers. Processing currently takes minutes per video, and the company has cited moving closer to real-time and exposing an API as roadmap items. A third-party profile lists AWS, React, Remix, React Router and Google Workspace in its technology stack.

Go-to-market

The company focuses exclusively on business customers, initially in professional training, marketing and media-related use cases, and has grown through named reference accounts such as TED, which publicly discussed results from using the technology at an industry conference. Funding proceeds were earmarked for product development across AI research, user experience and scalability rather than a specific sales expansion. Planned expansion areas include sports, education, marketing, healthcare and medicine, and an API is on the roadmap.

Business customers in media and entertainment, marketing and advertising, professional training, and creator/platform segments; named users and early adopters include TED Conferences and JFrog, with the company also citing the NBA. Stated expansion targets include sports, education, healthcare and medicine.

Geography

Third-party company profiles place Panjaya in Tel Aviv, Israel and in San Francisco, California; one of its co-founders holds the title of General Manager, Israel. Its product targets global audiences through localization into 29 languages.

Risks & controversies

The output of Panjaya's technology has been compared to deepfakes, and the company explicitly seeks to distance itself from that term and its connotations, describing itself as part of a "deep real" category. Management acknowledges misuse risk and cites B2B-only access, controlled customer onboarding and prospective watermarking and detection tooling as mitigations, stating an intent not to enable misinformation. A broader market risk noted in coverage is the widespread adoption of captions and subtitles as an alternative to dubbing. Third-party data aggregators report inconsistent basic facts about the company, including founding year (2021, 2022 or 2023), headquarters (Tel Aviv versus San Francisco) and total funding (with one estimate of $11 million against the $9.5 million reported in press coverage).