Flux

deepgram/flux

Flux is Deepgram's conversational speech recognition model for voice agents, launched October 2025, with turn detection inside the recognizer via StartOfTurn, EagerEndOfTurn and EndOfTurn events. Streaming-only on /v2/listen; this slug resolves to English flux-general-en.

Get API Key Read the Docs

import WebSocket from "ws";
import { readFile } from "node:fs/promises";

// Use mono, 16 kHz, signed 16-bit little-endian PCM.
const audio = await readFile("recording.pcm");
const url = new URL("/v1/audio/transcriptions/stream", process.env.AUDIOLAB_API_URL);
url.protocol = url.protocol === "https:" ? "wss:" : "ws:";
const socket = new WebSocket(url, {
  headers: { Authorization: `Bearer ${process.env.AUDIOLAB_API_KEY}` },
});
socket.on("open", () => socket.send(JSON.stringify({
  type: "session.start", model: "deepgram/flux",
  encoding: "pcm_s16le", sample_rate: 16000, channels: 1,
})));
socket.on("message", (data) => {
  const event = JSON.parse(data.toString());
  console.log(event);
  if (event.type === "session.started") {
    for (let offset = 0; offset < audio.length; offset += 3200)
      socket.send(audio.subarray(offset, offset + 3200));
    socket.send(JSON.stringify({ type: "session.finish" }));
  }
});

Input Price: $0.0065 / minute
Output Price: —
Input: Audio
Output: Text

Capabilities

Explore what Flux supports, then use the API section for a request tailored to this model.

StreamWord timestampsConfidence

Languages

English

Model Information

Provider: Deepgram
Product: Speech-to-Text
Release Date: Oct 2, 2025
Data Policy: Configurable
Region: Unknown
Input: Audio
Output: Text

Provider Documentation

Playground

Try Flux by Deepgram. Requests use the selected workspace’s credits and API pricing.

Flux

This model uses a dedicated input format. Explore its API example or open the supported playground.

Open playground

Providers

Compare available endpoints, pricing, and context limits for Flux.

Provider	Context	Output	Input Price	Output Price	Endpoint
Deepgram	Not specified	Text	$0.0065 / minute	—	`deepgram/flux:deepgram`

Prices are catalog rates. Your workspace’s estimate and endpoint pricing determine the final charge.

Input and Output Limits

Formats and limits AudioLab enforces on each request.

Limit	Value
Max bytes	Not specified
Max seconds	Not specified
Formats	flac, mp3, mp4, mpeg, mpga, m4a, ogg, wav, webm

Provider Notes

Nova 3 and Nova 2 support both uploaded audio and live streaming. Set diarize=true or provide diarize_model; AudioLab sends only diarize_model because Deepgram rejects requests that combine both flags. Flux General English and Flux General Multilingual are realtime-only conversational models; they use Flux end-of-turn controls and reject diarization. AudioLab maps common keyterms to Deepgram keyterm/keyword controls and rejects unsupported translation before dispatch.

Performance

No measurements in the last 24 hours.

Performance

Status

Current catalog availability. Your workspace’s access is checked separately before a request runs.

Checking availability…

A live availability check is not available right now.

Getting Started

Use your workspace API key and the exact model identifier below. Set the API URL to the endpoint shown in your workspace and keep credentials on your server.

import WebSocket from "ws";
import { readFile } from "node:fs/promises";

// Use mono, 16 kHz, signed 16-bit little-endian PCM.
const audio = await readFile("recording.pcm");
const url = new URL("/v1/audio/transcriptions/stream", process.env.AUDIOLAB_API_URL);
url.protocol = url.protocol === "https:" ? "wss:" : "ws:";
const socket = new WebSocket(url, {
  headers: { Authorization: `Bearer ${process.env.AUDIOLAB_API_KEY}` },
});
socket.on("open", () => socket.send(JSON.stringify({
  type: "session.start", model: "deepgram/flux",
  encoding: "pcm_s16le", sample_rate: 16000, channels: 1,
})));
socket.on("message", (data) => {
  const event = JSON.parse(data.toString());
  console.log(event);
  if (event.type === "session.started") {
    for (let offset = 0; offset < audio.length; offset += 3200)
      socket.send(audio.subarray(offset, offset + 3200));
    socket.send(JSON.stringify({ type: "session.finish" }));
  }
});

Read the Endpoint Reference

Request Parameters

Available parameters depend on the provider and request format. The endpoint reference explains accepted values and required fields.

Parameter	Details
`model`	`deepgram/flux` — Required model identifier.
`detect_language / detect_entities / diarize_model / multichannel`	View Accepted Values
`punctuate / smart_format / numerals / paragraphs / utterances / utt_split`	View Accepted Values
`filler_words / profanity_filter / dictation / measurements`	View Accepted Values
`intents / custom_intent / custom_intent_mode`	View Accepted Values
`topics / custom_topic / custom_topic_mode / summarize / sentiment`	View Accepted Values
`search / replace / tag / version / extra`	View Accepted Values
`endpointing / utterance_end_ms / vad_events / mip_opt_out (realtime)`	View Accepted Values
`eot_threshold / eager_eot_threshold / eot_timeout_ms / language_hints (Flux)`	View Accepted Values

Input and Output Limits

Formats and limits AudioLab enforces on each request.

Limit	Value
Max bytes	Not specified
Max seconds	Not specified
Formats	flac, mp3, mp4, mpeg, mpga, m4a, ogg, wav, webm

Similar Models

Explore more speech-to-text and compare their capabilities.

Model	Context	Input Price	Output Price
Flux MultilingualDeepgram	Not specified	$0.0078 / minute	—
flux-general-multiDeepgram	Not specified	$0.0066 / minute	—
flux-general-enDeepgram	Not specified	$0.0055 / minute	—
nova-3-medicalDeepgram	Not specified	$0.0037 / minute	—
Nova-3Deepgram	Not specified	$0.0037 / minute	—
nova-2-atcDeepgram	Not specified	$0.0037 / minute	—

View All Speech-to-Text