ComfyUI-EdgeTTS 2.0 adds SRT subtitles, targets voiceover failures
Version 2.0.0 of the third-party ComfyUI-EdgeTTS custom node adds an SRT subtitle output to its Whisper STT node and, according to its maintainer…
ComfyUI-EdgeTTS, a third-party custom node that brings Microsoft's Edge TTS voices into ComfyUI, adds a timecoded SRT output to its Whisper speech-to-text node in version 2.0.0. The maintainer, 1038lab, also says it addresses speech synthesis requests that had been failing with authorization errors. The project's update log labels the release 13 October 2026. GitHub records the release commit at 23:51:40 UTC on 13 October, which was 00:51:40 BST on 14 October in the UK.
For teams using ComfyUI for AI video production, the subtitle change is the more concrete addition. The update log says the Whisper STT node now has a dedicated SRT output socket next to its raw TEXT output. That SRT output produces subtitle blocks in the standard timecode format, such as 00:00:01,000 --> 00:00:03,500. The maintainer says creators need timecoded subtitles for video captioning and animation workflows. The SRT timestamps are formatted to the millisecond, but that formatting says nothing about how closely subtitle timings match the audio. The sources reviewed include no accuracy figures.
The 2.0.0 log describes the SRT output as having "zero external dependencies". An earlier entry in the same log, for v1.2.2 in January, added a note that Whisper STT requires FFmpeg, along with a Windows PATH setup example. Users should keep FFmpeg installed unless the maintainer states that the requirement has been dropped. The v1.2.2 entry also pins torchcodec 0.9 for torchaudio 2.9 and later.
The voiceover fix is the maintainer's account
According to the update log, upstream service protocol changes caused ComfyUI voiceover requests to fail with authorization rejection errors. The maintainer says v2.0.0 upgrades the node's internal connection negotiation. The maintainer also says existing workflows should keep working without reconfiguration. A retry step is meant to handle transient handshake failures during batch generation. Separately, the log describes audio decoding with automatic format fallbacks, intended to stop crashes caused by system codec mismatches. The sources reviewed contain no Microsoft statement about the protocol change. They do not show that access to the Edge TTS service will stay stable after this update.
The release also standardises output socket names to AUDIO, FILEPATH, TEXT and SRT. It adds tooltips to the Edge TTS, Save Audio and Whisper STT nodes and changes the module loader so nodes appear alphabetically in ComfyUI menus. The commit's file tree includes an example workflow that chains text-to-speech into speech-to-text. The README carries a GPL-3.0 licence badge.
Motion Engineering Today has not run the node. Voice generation, error recovery and SRT output were not tested. Before relying on 2.0.0 in a batch pipeline, teams should check three things on their own setup: that previously failing Edge TTS requests now complete, how the retry step behaves when the service rejects requests outright rather than briefly, and how closely SRT timings line up with the generated audio.
Frequently asked questions
What does ComfyUI-EdgeTTS 2.0 add for SRT subtitles?
ComfyUI-EdgeTTS 2.0.0 adds a dedicated SRT output socket to the node pack's Whisper speech-to-text node, next to its raw TEXT output. According to the update log, it produces subtitle blocks in standard timecode format, such as 00:00:01,000 --> 00:00:03,500. Millisecond formatting does not show how closely subtitle timings match the audio, and the sources reviewed include no accuracy figures. The update log labels the release 13 October 2026. Its GitHub commit is timestamped 23:51:40 UTC that day, or 00:51:40 BST on 14 October in the UK.
Does ComfyUI-EdgeTTS 2.0 fix Edge TTS voiceover authorization errors?
The maintainer, 1038lab, says upstream service protocol changes caused ComfyUI-EdgeTTS voiceover requests to fail with authorization errors. The maintainer says v2.0.0 upgrades the node's connection negotiation, adds a retry step for transient handshake failures during batch generation, and keeps existing workflows working without reconfiguration. Motion Engineering Today has not independently tested these claims. The sources reviewed contain no Microsoft statement about the protocol change and do not show that Edge TTS access will stay stable. Teams should confirm that previously failing requests now complete on their own setup.
Do I still need FFmpeg for Whisper in ComfyUI-EdgeTTS 2.0?
Keep FFmpeg installed unless the maintainer says the requirement has been dropped. The ComfyUI-EdgeTTS 2.0.0 update log describes the SRT output as having "zero external dependencies". The v1.2.2 entry in the same log, from January, states that Whisper STT requires FFmpeg and gives a Windows PATH setup example.
AI-assisted, reviewed by Codex (25-post review authorized by JJ Rooney, 14 October 2026) on 14 October 2026.
