Hi team, running into an error when setting Gladia as the transcriber. 1. Set Gladia as the transcriber provider 2. Set Solaria as the model 3. Publish and talk to assistant The call fails to initialize, the following error is seen in the logs: { "id": "log-13", "level": 50, "time": 1772145240224, "timestamp": 1772145240224000000, "severityText": "ERROR", "severityNumber": 17, "body": "Gladia V2 initialization failed", "attributes": { "category": "transcriber", "callId": "019c9c16-6460-755e-a304-xxxxxxxxxxx", "orgId": "84accdd9-fe2e-45bb-992c-xxxxxxxxxxxxxx", "provider": "gladia", "type": "assistant", "error": { "message": { "message": "Request failed with status code 400", "data": { "statusCode": 400, "timestamp": "2026-02-26T22:34:00.145Z", "path": "/v2/live", "message": "Invalid parameter(s). See validation_errors for more details.", "validation_errors": [ "sample_rate must be one of the following values: 8000, 16000, 32000, 44100, 48000" ], "request_id": "G-4b42fa97" } }, "code": "ERR_BAD_REQUEST", "status": 400, "statusText": "Bad Request" } } }

To resolve the Gladia V2 initialization error, set your

sample_rate

to a supported value (e.g.,

16000

). Update your configuration as follows:

json
{
  "provider": "gladia",
  "model": "solaria",
  "sample_rate": 16000
}

Save and redeploy your assistant after making this change. For more details, see the [Gladia transcriber documentation](https://docs.vapi.ai/providers/transcriber/gladia). Source: - [Gladia transcriber documentation](https://docs.vapi.ai/providers/transcriber/gladia)

@User sample_rate is not a supported configuration for the transcriber: "transcriber.property sample_rate should not exist"

Hi, Please try these steps so we can pinpoint the issue quickly: 1. Verify API key & plan • Confirm the Gladia API key is active and not rate/usage limited (402 if free minutes exhausted). See the error pattern in related threads. • If unsure, generate a new key in your Gladia account and paste it into Vapi. 2. Test in Dashboard • Open your assistant > Transcriber tab > Provider = Gladia and click "Talk to Assistant" to test (removes telephony variables). See docs: [https://docs.vapi.ai/providers/transcriber/gladia](https://docs.vapi.ai/providers/transcriber/gladia) 3. Check Observe logs • Go to Observe > Call Logs and API Logs. Copy any call ID and the exact error (e.g., "gladia-transcriber-failed" or 402/invalid credentials). 4. Confirm transcriber config • Ensure language(s) and model are set correctly (e.g., set language to the target language or use automatic detection). For custom vocabulary or advanced settings, check changelog options if needed: [https://docs.vapi.ai/changelog/2025/8/3](https://docs.vapi.ai/changelog/2025/8/3) 5. Try fallback • Configure a fallback transcriber/provider in the assistant in case Gladia fails. If you still have issues, let us know. We will investigate further.