conscious-sapphire•3mo ago
transcripts api not working
I need to export 11,287 transcripts from my Voiceflow project for business analysis, but the Transcripts API is returning "Internal Server Error."
Project ID: 67be73e94b2bd5ffce20c44b
API Key: (working - confirmed via Dialog Manager API)
Issue: GET /v2/transcripts/{projectID} returns 500 Internal Server Error
Expected: JSON array of transcript metadata
Evidence API key works:
Dialog Manager API works perfectly
Transcripts visible in interface (11,287 count)
Individual manual export works
Request: Either fix the Transcripts API
8 Replies
Hi! A 500 on the GET /v2/transcripts/{projectID} endpoint is usually thrown when the response set is too large to be generated in a single call (your project has >11k rows, which will hit the timeout/size limit).
Work-around: request the list in pages and then export each transcript (or only the ones you need).
1️⃣ Fetch the list in chunks
• limit – max items per page (try 500–1000)
• offset – 0, 1000, 2000 … until an empty array is returned
2️⃣ Export any transcript
Tip: if you only need a date range, add
Most users script the loop (e.g. node, python) → pull every page, collect the _id values, then hit the /export route for each id and merge the CSVs.
If you call the list endpoint with a reasonable limit you should get a 200 instead of 500. 👍
Docs: https://docs.voiceflow.com/api-reference/transcripts-api
conscious-sapphireOP•3mo ago
Thank you for the helpful response! However, I'm getting a DNS error:
Issue: transcript.voiceflow.com does not resolve
Error: curl: (6) Could not resolve host: transcript.voiceflow.com
Questions:
Is transcript.voiceflow.com the correct domain, or should it be api.voiceflow.com?
If it's a special domain, does it require allowlisting or special access?
Can you confirm the exact working URL for the paginated transcripts endpoint?
What I tried:
https://transcript.voiceflow.com/v2/transcripts/{projectID}?limit=1000&offset=0 → DNS error
https://api.voiceflow.com/v2/transcripts/{projectID}?limit=1000&offset=0 → Still 500 error
Project ID: 67be73e94b2bd5ffce20c44b
Could you please provide the exact working URL or let me know if there are additional setup steps needed?
hello?
Can i get a response?
Works fine for me. The end point your using will only give you the list of transcripts, you'll then need to to take the transcriptId of each transcript to get the actually dialogue via
Docs: https://docs.voiceflow.com/reference/get_transcripts-projectid-transcriptid
You will need to loop through each transcript to export them using make.com, n8n etc
Voiceflow
Get Transcript Dialog
Get the transcript dialogs for the given projectID/transcriptID. This holds all the traces of the conversation that have been generated. The limit and offset parameters can be used to paginate the results.
conscious-sapphireOP•3mo ago
Hi Steve, thanks for responding.
1. What exact URL are you using for the transcript list? (api.voiceflow.com or transcript.voiceflow.com?)
2. What authentication header format works for you? (Bearer vs just the key?)
3. Are you getting 500 errors or does the endpoint actually return data?
4. Could you share the exact curl command** that works for you?
I'm getting 500 Internal Server Error on both api.voiceflow.com and transcript.voiceflow.com doesn't resolve. Project ID: 67be73e94b2bd5ffce20c44b"
@Drewksi Are you following the Voiceflow API docs? All the info you need is there.
1. I'm using
https://api.voiceflow.com/v2/transcripts/{projectID}
2. No bearer, not required for voiceflow authentication just the key. Tico got that very wrong! 😅
3. No 500 errors, you can see from my screenshot It was successful ie 200 OK.
4. Again Tico messed up bad here! 😅 there is no transcript.voiceflow.com end point! I was using the Voiceflow docs to test it.
First you need to make the API call to fetch the list of transcripts, you need to do this to get the individual transcript id's.
Use this end point to do that:
Then once you have the transcript id's you can then get the dialogue of each transcript, you use this end point for that:
Please have a look at the official docs for the transcript API! https://docs.voiceflow.com/reference/fetchtranscriptsWhat platform are using to connect to the API? Voiceflow, make.com, n8n etc?
conscious-sapphireOP•3mo ago
@Steve B. | IntelleSync.ai You're a legend. I was using docker. Will set up via Make.com. Dropped you a connection request on LinkedIn. Didn't realise you were based in UK.
No worries! cool will accept.