Every reverse engineering of the Google Speech API requires an API key but Chrome is able to call the server, seemingly without one. How does this work internally?
Is it possible to use the API for any sort of large scale speech transcription?
Asked
Active
Viewed 1,523 times
1 Answers
0
Looking into the Chromium source code, http://src.chromium.org/viewvc/chrome/trunk/src/content/browser/speech/google_one_shot_remote_engine.cc indicates that the server is passed an API key with the request.
It also seems that Google Chrome comes with an API key, and chromium can, depending on the distribution. https://code.google.com/p/chromium/wiki/ChromiumBrowserVsGoogleChrome
It's still unclear to me why the browser calls to the server are not effected by the 50 calls/day limit.
knightian
- 683
- 1
- 7
- 20