Issue: I chosed gladia as transcriber and it cannot work in assistant and i tried api key integration and it says api key wrong. 1. Navigate to assistant and select an assistant 2. Change transcriber to gladia and model solaria. Select all languages 3. Publish and talk to assistant Error: the transcription does not work and ai cannot hear the human https://www.loom.com/share/3f41a265675a44b7b4c478d3ae10e9df?sid=d7a644f9-7f5e-4d36-b3de-cca1756f04ed

Suggestion: Double-check that your Gladia API key is correct and properly entered in the Vapi dashboard. Go to Settings > Provider Keys, remove any existing Gladia key, and re-add your API key—make sure there are no extra spaces or hidden characters.

text
# Example: Enter your Gladia API key exactly as provided by Gladia
GLADIA_API_KEY=your_actual_gladia_api_key

After saving, try publishing and testing your assistant again. If the error persists, verify your Gladia account status or try generating a new API key from the Gladia dashboard. Sources: - [Gladia Transcription Provider Support Vapi Docs](https://docs.vapi.ai/changelog) - [Gladia Provider Overview](https://docs.vapi.ai/providers/transcriber/gladia)

i think there is an issue with my account, i cannot remove members also. and i tried all types of integration cannot add api keys

Hi SamSEOClub, Based on your issue with Gladia transcriber, here are a few things to check: 1. From the documentation about Gladia transcriber: - Gladia is an advanced AI platform for real-time transcription that supports 99 languages - Make sure you have set up and verified your Gladia API key properly 2. For configuration: 3. Go to "Assistants" tab in the Vapi dashboard 4. Click on your assistant and select the "Transcriber" tab 5. Select "gladia" as the Provider 6. Set the model to "solaria" 7. Make sure you have added your Gladia API key in the credentials/settings 8. Click "Publish" to save changes If the AI still cannot hear the human: 1. Check that your microphone permissions are enabled 2. Verify the Gladia API key is valid and properly configured 3. Try testing with a different transcriber provider (like Google) to isolate if it's a Gladia-specific issue If the problem persists after verifying these steps, I recommend: 1. Double-checking your API key is correctly entered 2. Trying a different transcriber provider temporarily to help troubleshoot 3. Checking the call logs for any specific error messages

I have done checking all of the above and infact i cannot insert any api keys sucessfully, meaning it will say im having error. Are u able to investigate my account??

Hi SamSEOClub, Have you tried to set up AssemblyAI as your transcriber, follow these steps: 1. Go to the "Assistants" tab in your Vapi dashboard. 2. Click on your assistant and then select the "Transcriber" tab. 3. Choose "assembly-ai" from the Provider dropdown. For more details, you can refer to the [AssemblyAI documentation](https://docs.vapi.ai/providers/transcriber/assembly-ai). Note that AssemblyAI's Universal-Streaming model currently supports English only. If you need support for other languages, consider using a different provider like Google, Talkscriber, or Azure, which offer more language options as detailed in the documentation.

can u help to check my account??? today suddenly minimax cannot work and all the integration i cannot insert api key! says error i already copied correct!

Of course, can you give your ORG ID?

Yes

We reviewed the credential that is stored in the database and it is a JWT token instead of an API key. Please use the API key from the MiniMax dashboard and it should work. Also remove your current one as it is not being accepted, which is causing the insufficient token error

why is the multi lingual transcriber for deepgram and speechmatics not working?

Sounds good. Let us know if you need anything else

You hit your rate limit on your minimax credits and no fallback was configured, which made the call end early:

🟡 06:09:34:850 Minimax (WebSocket Pool) Error: {
  "message": "rate limit",
  "stack": "Error: rate limit\n    at MinimaxWebsocket.messageHandle (/app/dist/libs/core/src/components/liveCallPipeline/streams/synthesizers/minimaxStream.js:274:32)\n    at MinimaxWebsocket.socket.onmessage (/app/dist/libs/core/src/components/liveCallPipeline/streams/synthesizers/minimaxStream.js:82:22)\n    at callListener (/app/node_modules/ws/lib/event-target.js:290:14)\n    at WebSocket.onMessage (/app/node_modules/ws/lib/event-target.js:209:9)\n    at WebSocket.emit (node:events:520:28)\n    at WebSocket.emit (node:domain:488:12)\n    at Receiver.receiverOnMessage (/app/node_modules/ws/lib/websocket.js:1220:20)\n    at Receiver.emit (node:events:520:28)\n    at Receiver.emit (node:domain:488:12)\n    at Receiver.dataMessage (/app/node_modules/ws/lib/receiver.js:596:14)\n    at Receiver.getData (/app/node_modules/ws/lib/receiver.js:496:10)\n    at Receiver.startLoop (/app/node_modules/ws/lib/receiver.js:167:16)\n    at Receiver._write (/app/node_modules/ws/lib/receiver.js:94:10)\n    at writeOrBuffer (node:internal/streams/writable:570:12)\n    at _write (node:internal/streams/writable:499:10)\n    at Writable.write (node:internal/streams/writable:508:10)"
}
🔵 06:09:34:859 TTS stream error but no fallback plan configured or no fallback voices remaining. Failing with error pipeline-error-minimax-voice-failed.

To prevent this, configure a fallback in case your minimax rate limit is hit. To prevent rate limiting, remove your minimax API key and just use VAPI default API key for minimax.

The lag on calls... anyway to solve it??? Using 4.1 needs it for mcp and better processing thoughts and more natural.