HD Group Audio Chat

The WebSocket version turns any conversation, group chat, or chat room into a multi-party voice call supporting up to 50 simultaneous participants. The feature uses WebRTC with a dedicated media server for efficient routing — call quality stays high even with dozens on the line, while bandwidth stays modest compared to video.

WebSocket version

Group audio chat is a WebSocket-version feature. The free version does not include voice calling.

What it adds#

Up to 50 simultaneous voice participants per call
Works in private one-on-one conversations, group chats, and chat rooms
Lower bandwidth than group video — typically 50–80 kbps per participant
Media-server routing (SFU model) for stable quality at scale
Remote mute/unmute control for the call host
Per-role restrictions: control which user roles can place or join calls
Optional mini-chat call buttons in members directories and on profile pages
Bandwidth-friendly fallback for users on cellular / low-quality connections

How it works#

A user clicks the Voice Call button in a conversation. Better Messages signals the request through the WebSocket relay. The cloud server allocates a call room on the media server. All participants connect to the media server via WebRTC. Each participant uploads their own microphone stream once. The media server distributes streams to everyone else — this is dramatically more efficient than full mesh (where each participant uploads to every other participant individually).

Topology	Network upload per participant
Full mesh (peer-to-peer)	(N−1) × stream — quadratic, breaks past 5–6 participants
Media server (SFU)	1 × stream — flat, scales to 50+

The host (the user who started the call) can remotely mute disruptive participants. Participants can also self-mute, self-unmute, and leave anytime.

The call session is independent of the underlying conversation thread — when the call ends, the conversation remains exactly as it was, with a system message noting "Voice call ended" and duration.

When to use group voice over group video#

Choose group voice when…	Choose group video when…
10+ participants	Visual presence matters
Mobile/cellular users	High-bandwidth connections
Discussion / audio-first content	Demo, screen-share, face-to-face
Bandwidth-constrained users	Standard broadband
Privacy preference for audio only	Trust-building, sales / coaching

Requirements#

Site must be served over HTTPS (browser WebRTC requirement)
Each participant needs a working microphone
Modern browser with WebRTC support (Chrome, Firefox, Edge, Safari, mobile equivalents)
WebSocket-version license active

Configuration#

Per-role restrictions and optional mini-chat call buttons apply to both audio and video group calls — see Group Video Chat → Per-role restrictions for the shared admin controls.

Frequently asked questions#

What is the maximum number of participants?#

50 simultaneous voice participants. Sites that regularly need more can contact support to discuss the self-hosted plan, where the cap can be raised based on media-server capacity.

Are group voice calls recorded?#

No — calls are real-time only and not recorded by default. The plugin doesn't include built-in recording. Sites that need recording typically use a screen-recording layer at the participant level or browser-based recording extensions, subject to participant consent.

Can guests (non-logged-in users) join a group voice call?#

Yes — if guest chat is enabled on the chat room, guests can join voice calls placed in that room. They appear by their entered display name.

What's the bandwidth requirement per participant?#

About 50–80 kbps upload + 50–80 kbps download per active speaker. A 20-person call uses about 1.5 Mbps download on each device. The codec is Opus.

Does the media server store anything?#

No. The media server routes audio streams in memory only — no storage, no recording, no transcription on the audio path. Call metadata (start time, participants, duration) is logged for the host's call history.