Adobe and Speechmatics have announced a significant update to their long-standing partnership, introducing a new on-device speech-to-text (STT) model for Adobe Premiere Pro. This development allows creators to generate high-accuracy transcripts locally on their hardware, eliminating the need for an active internet connection or cloud processing. By bringing cloud-grade performance to the edge, the integration addresses the growing demand for data privacy and production flexibility in the 2026 media environment.
The partnership between these two companies dates back to 2021, when Premiere Pro became the first major non-linear editing platform to integrate Speechmatics’ transcription technology. The latest evolution of this collaboration focuses on closing the performance gap between local and cloud-based AI. The new model is optimized to run efficiently on modern processors, providing studios, agencies, and independent creators with the same level of precision they would expect from a remote server while keeping all audio data secure on the local device.
For podcast creators and video editors, the move to on-device processing offers practical advantages beyond security. Editors can now maintain their workflows in environments where connectivity is limited or non-existent, such as on film sets, during travel, or in high-security facilities with restricted network access. This reduces friction in the content supply chain, as there is no longer a delay caused by uploading large audio files for transcription.
The technical improvements in this update specifically target real-world audio challenges, including diverse accents and background noise. Speechmatics’ models are designed to handle "messy" audio, which is common in field recordings and remote interviews. By integrating these capabilities directly into the Premiere Pro interface, Adobe continues to streamline the editing process, allowing users to move from raw footage to a searchable, text-based edit in a single environment.
This update also aligns with broader industry trends toward sovereign AI and data governance. Large enterprises and production houses often face strict compliance requirements that prohibit the use of cloud-based transcription for sensitive or unreleased content. On-device processing ensures that intellectual property remains within the user's controlled infrastructure, providing a layer of protection that is increasingly necessary as AI becomes more central to the creative process.
Article on the best AI transcription tools for enterprise workflows
Beyond the privacy benefits, the on-device model supports the growing use of agentic AI and natural language interfaces within creative software. As speech becomes a primary way for users to interact with their projects—such as using voice commands to find specific moments in a recording—the foundational accuracy of the transcription layer becomes critical. The Speechmatics integration ensures that these advanced features work reliably across 55+ languages, supporting a global community of storytellers.
Creators can access the new on-device capabilities through the latest version of Adobe Premiere Pro. As the industry moves further toward decentralized and mobile production, the ability to process complex audio data locally will likely become a standard requirement for professional media tools. For businesses looking to scale their video and audio output, these advancements represent a significant step toward more efficient, secure, and accessible storytelling workflows.
Details on the rollout and technical requirements are available through the official Adobe and Speechmatics partnership documentation.
