From e025f565b213cf42d6cfdef15087e93a24a391aa Mon Sep 17 00:00:00 2001
From: Stephen Smith <stephen@vapi.ai>
Date: Fri, 17 Jul 2026 13:38:01 -0700
Subject: [PATCH] Added Sonix transcriber

---
 .../transcriber-fallback-plan.mdx             | 11 +++-
 fern/docs.yml                                 |  2 +
 fern/providers/transcriber/soniox.mdx         | 54 +++++++++++++++++++
 3 files changed, 66 insertions(+), 1 deletion(-)
 create mode 100644 fern/providers/transcriber/soniox.mdx

diff --git a/fern/customization/transcriber-fallback-plan.mdx b/fern/customization/transcriber-fallback-plan.mdx
index 4c1b4db88..34b8b71c8 100644
--- a/fern/customization/transcriber-fallback-plan.mdx
+++ b/fern/customization/transcriber-fallback-plan.mdx
@@ -171,6 +171,15 @@ Each transcriber provider supports different configuration options. Expand the a
     - **enableDiarization** (boolean): Enable speaker identification for multi-speaker conversations.
     - **maxDelayMs**: Maximum delay in milliseconds for partial transcripts. Balances latency and accuracy.
   </Accordion>
+  <Accordion title="Soniox">
+    See [Configure Soniox](/providers/transcriber/soniox) for setup instructions.
+
+    - **model**: Model selection (currently only `stt-rt-v5` for new configurations).
+    - **language**: ISO 639-1 language code. Soniox supports multilingual transcription with a single universal model.
+    - **languageHintsStrict** (boolean): Restrict transcription to the selected language when enabled. When disabled, Soniox can automatically detect and transcribe supported languages. Default is `true`.
+    - **maxEndpointDelayMs** (500-3000): Maximum delay in milliseconds between the end of speech and endpoint detection. Lower values provide faster turn-taking but may increase false endpoints. Default is `500`.
+    - **customVocabulary**: Array of brand names, product names, or other domain-specific terms to improve recognition accuracy.
+  </Accordion>
   <Accordion title="Google">
     - **model**: Gemini model selection.
     - **language**: Language selection (e.g., `Multilingual`, `English`, `Spanish`, `French`).
@@ -205,7 +214,7 @@ Each transcriber provider supports different configuration options. Expand the a
     Auto fallback lets Vapi intelligently select an alternative provider for you—no configuration needed. Manual fallback lets you specify exact providers in a specific priority order. You can use both together: manual fallbacks are tried first, and auto fallback acts as a safety net if they all fail.
   </Accordion>
   <Accordion title="Which providers support fallback?">
-    All major transcriber providers are supported: Deepgram, AssemblyAI, Azure, Gladia, Google, Speechmatics, Cartesia, ElevenLabs, and OpenAI.
+    All major transcriber providers are supported: Deepgram, AssemblyAI, Azure, Gladia, Google, Speechmatics, Soniox, Cartesia, ElevenLabs, and OpenAI.
   </Accordion>
   <Accordion title="Does fallback affect pricing?">
     No additional fees for using fallback transcribers. You are only billed for the transcriber that processes the audio.
diff --git a/fern/docs.yml b/fern/docs.yml
index 31f77c4b4..8001d7a02 100644
--- a/fern/docs.yml
+++ b/fern/docs.yml
@@ -667,6 +667,8 @@ navigation:
                     path: providers/transcriber/gladia.mdx
                   - page: Speechmatics
                     path: providers/transcriber/speechmatics.mdx
+                  - page: Soniox
+                    path: providers/transcriber/soniox.mdx
                   - page: Talkscriber
                     path: providers/transcriber/talkscriber.mdx
                   - page: AssemblyAI
diff --git a/fern/providers/transcriber/soniox.mdx b/fern/providers/transcriber/soniox.mdx
new file mode 100644
index 000000000..b0bb32cfb
--- /dev/null
+++ b/fern/providers/transcriber/soniox.mdx
@@ -0,0 +1,54 @@
+---
+title: Soniox
+subtitle: Configure real-time, multilingual speech-to-text for Vapi assistants.
+description: Connect Soniox to a Vapi assistant and configure the stt-rt-v5 real-time speech-to-text model for low-latency, multilingual transcription and endpointing.
+slug: providers/transcriber/soniox
+---
+
+Soniox is a real-time speech-to-text (STT) provider that transcribes caller audio for Vapi assistants. Use the `stt-rt-v5` model for new configurations.
+
+## What is Soniox?
+
+Soniox provides low-latency, multilingual transcription and semantic endpointing for voice assistants. The `stt-rt-v5` model supports real-time audio and is the current model for new Soniox configurations.
+
+## Connect Soniox
+
+Configure Soniox in the Vapi Dashboard or update an assistant through the API.
+
+<Tabs>
+  <Tab title="Dashboard">
+    <Steps>
+      <Step title="Open an assistant">
+        In the [Vapi Dashboard](https://dashboard.vapi.ai), open **Assistants**, then select the assistant you want to update.
+      </Step>
+      <Step title="Select Soniox">
+        Open the **Transcriber** settings. Set **Provider** to **Soniox** and **Model** to **stt-rt-v5**.
+      </Step>
+      <Step title="Publish the assistant">
+        Click **Publish** to apply the new transcriber configuration.
+      </Step>
+    </Steps>
+  </Tab>
+  <Tab title="cURL">
+    Set `VAPI_API_KEY` in your environment, then replace `YOUR_ASSISTANT_ID` with the assistant ID.
+
+    ```bash
+    curl -X PATCH "https://api.vapi.ai/assistant/YOUR_ASSISTANT_ID" \
+      -H "Authorization: Bearer $VAPI_API_KEY" \
+      -H "Content-Type: application/json" \
+      -d '{
+        "transcriber": {
+          "provider": "soniox",
+          "model": "stt-rt-v5",
+          "language": "en"
+        }
+      }'
+    ```
+  </Tab>
+</Tabs>
+
+## Related documentation
+
+- Review the [`SonioxTranscriber` fields](/api-reference/assistants/update) in the Update Assistant API reference.
+- Configure a [transcriber fallback plan](/customization/transcriber-fallback-plan).
+- Review the [Soniox speech-to-text models](https://soniox.com/docs/stt/models).