Google Speech API claims to support a number of codecs (https://cloud.google.com/speech/docs/basics). I'm interested in processing an archive of a session produced by TokBox WebRTC.
Is there a sample code that does something like this? Does the archive need to be converted to a compatible format?
The default audio codec for WebRTC is Opus, which is indeed supported by Google Speech API. The trick is getting the audio out of an OpenTok stream and forwarded along to the recognition service; unfortunately this is no small effort.
Although some work has been done on this in an experimental capacity, there is no official support at this time. Recommend reaching out to TokBox support directly to discuss the specifics of what you're trying to build (email support at tokbox.com).
Disclosure: I work at TokBox.