You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Published npm 4.43.3 still missing assemblyai-stt-3-7 realtime models in StreamingSpeechModel #216
On the published package assemblyai@4.43.3 from npm, StreamingSpeechModel does not include either value. TypeScript projects following the launch quickstart fail under --strict.
There is already an internal PR for this (#210), but it is still open / dirty and has not landed on npm. From a developer perspective the launch shipped without the JS types.
• Python SDK 1.10.11 already includes both enum values.
• StreamingMode on npm also lacks "manual", while docs say realtime-agent supports it and realtime rejects it with code 3006.
• Related: #215 covers pre-recorded language_codes typing for STT-3.7
Hey @webski101, thanks for pointing this out, this was a miss on my end. The PR in this repo was never merged by mistake. You can find the realtime 3-7 models in the newly published v4.43.4
Activity
ccampbell-aai commentedon Nov 19, 2026
Hey @webski101, thanks for pointing this out, this was a miss on my end. The PR in this repo was never merged by mistake. You can find the realtime 3-7 models in the newly published
v4.43.4