A voice adds another layer to the conversation

Alys can display a reply as text and, when voice is enabled, speak it through the character in the scene. Lip-sync animation accompanies the audio. This makes a typed conversation feel more present without requiring you to speak into a microphone.

Voice is optional. You can keep a quiet conversation in chat bubbles, enable sound when it suits you and return to text whenever you prefer.

Begin with a short sample

Open the voice section in Your Alys and select a voice to hear a brief sample. The preset collection contains ten voices, described by accent and tone rather than by a second character name. You are choosing how Alys sounds.

Listen for qualities you would enjoy over several replies: clarity, pace, warmth and how naturally the voice handles a question. A dramatic sample can be fun, but a gentler delivery may be more comfortable for longer conversations.

Save the voice you prefer. Audio begins through a user interaction, so do not expect a newly opened page to start speaking before you enable it. Device volume and browser audio permissions still apply.

Choose from the voice catalogue

Paid members can choose from ten catalogue voices and hear a short sample. Samples and spoken replies consume the displayed spoken-character allowance. Custom voice IDs are not supported. Text remains available within its chat allowance when voice runs out.

What happens during a reply

Text begins appearing while the reply is generated. Speech follows in segments, with mouth animation tied to playback. The experience depends on the generation service, speech service, connection and device; it is not a guarantee that every response starts at the same speed.

Stopping a reply should stop its speech as well. Leaving the page or switching away should not leave old audio playing unexpectedly. If sound fails, the text conversation remains the useful fallback.

For a first test, ask for a short answer. It is easier to judge pronunciation and timing with two sentences than with a long monologue.

Current limits and supported input

Alys currently takes typed messages. Spoken output does not mean the app is listening to you, and this version does not offer microphone conversation or a live video call.

Free includes 50 text replies each month. Casual is $6.99/month for 500 replies and 20,000 spoken characters; Committed is $12.99/month for 2,000 replies and 50,000 spoken characters. Prices are in USD. See plans and allowances for current availability. Spoken characters measure the text sent for speech, including spaces and punctuation. For example, 20,000 characters is roughly 40 replies of 500 characters each; actual reply length varies, so this is not a promised number of voice replies. Previews also consume characters. Monthly allowances are used before prepaid packs.

Changing the visual expression or camera is also separate from the words being generated. Lip sync animates speech; it is not evidence that Alys sees the room around you.

Choose the experience that fits the moment

Use a compact chat panel when you want to watch the character, or expand the chat when reading matters more. Turn voice off in shared spaces and use headphones when you want a more personal experience.

Meet Alys in the browser, read getting started with Alys, or compare the difference between spoken replies and broader calling features in Alys vs Kindroid.

Feature guide from the Alys team, updated 6 October 2026.