What's new
- Text to speech: the local buyer supports the OpenAI speech API (
POST /v1/audio/speech). - Image edits and lip-synced video: the
antseed-imagesandantseed-videosskills can edit images you bring and animate a character to an audio track. - Payments: fixes for transactions stuck on a low fee and for stored channels the chain shows as closed.
- Model verification results now come from the current verification contracts on Base.
- Fewer buyer memory problems, and a fix for a race when watching channels early in a run.
