Guide · Flowise

Saidwire speech-to-text in Flowise

Flowise turns voice messages into text with the speech-to-text provider you choose for each chatflow. Its LocalAi STT provider calls any server that is compatible with OpenAI's transcription API and has a Base URL field. Point that field at Saidwire and voice input in the Flowise chat and the embedded chat widget runs on Saidwire. You don't need to write any code.

you need
Flowise	Cloud or self-hosted, a recent version with Configuration → Speech to Text. Checked against the Flowise source as of August 2026.
Saidwire API key	`sk_live_...` from saidwire.com/dev (120 free hours, no card)

1. Open the speech-to-text settings

Open your chatflow or agentflow.
Click the settings (gear) icon at the top right of the canvas, then choose Configuration.
Under Media & Files, open Speech to Text.

2. Choose LocalAi STT and fill it in

Under Providers, choose LocalAi STT.
Under Connect Credential, choose Create New. Name the credential Saidwire and paste your key into LocalAI Api Key, then add it.

Fill in the rest:

field	value
Base URL	`https://api.saidwire.com/v1`
Language	`en`, or leave it empty
Model	`saidwire-en`. If you leave it empty, Flowise sends `whisper-1`, which Saidwire also accepts.
Prompt	leave it empty (Saidwire ignores it)
Temperature	leave it empty (Saidwire ignores it)

Click Save. The button stays greyed out until a credential is selected.

3. Try it

Open the chat for the flow, click the microphone, say a sentence and send it. Flowise shows what you said as your message and passes the text on to the flow. If the key is wrong, the error appears on that first voice message, because Flowise doesn't test this credential when you save it.

What Flowise sends

For each voice message, Flowise uploads the recording to POST https://api.saidwire.com/v1/audio/transcriptions with your key as a Bearer token. The recording is WebM from Chrome and Edge, or MP4 from Safari. Flowise also sends model, and sends language, prompt and temperature when you set them. It reads back the text field. We sent this exact request, using the same client library versions Flowise uses, against the live API before publishing this page.

Use LocalAi STT rather than OpenAI Whisper. The OpenAI Whisper option has no Base URL field. You could redirect it with the server-wide OPENAI_BASE_URL environment variable, but that also sends every other OpenAI call in your Flowise to Saidwire.

Troubleshooting

you see	fix
`401 Incorrect API key provided`	The key in the LocalAI API credential is wrong, has been revoked, or has a stray space.
`404`	Set Base URL to exactly `https://api.saidwire.com/v1`. Flowise adds `/audio/transcriptions` itself.
`400 ... unsupported_language`	Saidwire transcribes English only. Set Language to `en` or leave it empty.
`429`	Too many voice messages at once for your key (trial keys allow 4), or your free hours are used up. The message says which.

Good to know

Each flow keeps its own speech-to-text setting, so repeat these steps for every flow that should use Saidwire.
This setup gives you text only. If you need word timestamps or SRT/VTT subtitles, call the endpoint directly. The API docs list every field.
Files can be up to 100 MB and 60 minutes. WAV, MP3, FLAC, Ogg/Opus, WebM and M4A/MP4 work. For M4A/MP4 files, the index has to be at the start of the file. Browser recordings already are.
Saidwire processes the audio in memory and does not store it.
After the free hours, Saidwire costs $0.30 per audio hour, billed per second.

Stuck? Email support@saidwire.com with the x-request-id from the failed call.

OpenAI is a trademark of OpenAI. Open WebUI, n8n, LiteLLM, Dify, Flowise, Vercel and the AI SDK belong to their respective owners. Saidwire is not affiliated with OpenAI or with these projects.