Skip to content
VALA Worlds

Call for voices · Voice actors

We are looking for four voices.

For VALA, our narrative game, we are looking for two female and two male voices of audiobook or audio drama quality. From your recordings we will build AI voices for VALA, with your express consent and in return for payment.

A note before you read on: the recordings are in German. The application documents and the contract are in German too. This page is a translation for your convenience; should the two versions differ, the German version prevails.

Roles
Four voices, each for narration and characters
Recording
About 60 minutes from a script
Use
For VALA only, non-exclusive
Applications
By 7 December 2026

VALA Worlds is a small family business from Pinneberg in Schleswig-Holstein, Germany, founded in 2026 and funded from our own resources. We are looking for voices for VALA, our narrative game, which is played not on a screen but in conversation.

A voice tells you what is happening in a world. You answer in your own words, the characters react, and the world remembers what has happened.

We want to bring back what screens so often crowd out: imagination, listening and telling stories together, alone in an armchair or with others around the table.

That is why, in VALA, the voice is everything. Because the story only comes into being in conversation, it cannot be recorded in a studio beforehand. We are looking for four voices from which, with your express consent and in return for payment, we will build AI voices for VALA.

VALA has several worlds, among them Restoration-era Paris, ancient Rome, the fantasy world Sage, the science-fiction world Helios and the thriller Anvertraut. More worlds will follow. The voices we are looking for are to be used in all present and future VALA worlds.

Wanted: four voices

We are looking for two female and two male voices of audiobook or audio drama quality, one younger and one older of each. Each voice narrates and plays characters.

RoleVocal age, approx.Use
Woman, younger25–40 yearsNarrator and characters
Woman, older50–70 yearsNarrator and characters
Man, younger25–40 yearsNarrator and characters
Man, older50–70 yearsNarrator and characters

Only the sound of the voice counts. Your actual age and gender play no part in the selection.

What we hope for

  • experience in audiobooks or audio drama
  • an assured switch between a calm narrative tone and lively character speech
  • standard German (Hochdeutsch) without a strong regional colouring
  • confident pronunciation of French names and places

You may apply for one or two roles.

What will be recorded

About 60 minutes of finished material per voice, from a script that we supply.

  • calm narration, about 20 minutes
  • five emotions as character speech, about 6–8 minutes each: happy, sad, angry, tense to whispered, tender
  • lines of dialogue, numbers and dates
  • French, English and Latin names and places, with a pronunciation list

Recording takes place in your own studio or a studio of your choice. Studio costs and one round of pick-ups (up to 10 minutes) are to be included in your quote.

Delivery format

  • WAV, 48 kHz, 24-bit, mono
  • dry: no reverb, no music, no compression, no limiter, no noise reduction
  • everything in the same room, with the same microphone at the same distance
  • one file per script line, mostly 5–12 seconds long; file name = line number
  • levels: peaks between −12 and −3 dBFS, nothing clipped
  • room noise floor below −60 dBFS; no filter other than a low-cut (rumble) filter up to 80 Hz
  • about half a second of silence before and after each line; breaths stay in
  • plus, once, 30 seconds of room tone, that is, silence in the room
  • if the spoken text departs from the script, please re-record or note the deviation

What the voice will speak

The AI voice speaks texts that come into being during play. Nobody can approve every sentence in advance, and neither can we.

How the technology works

We use a freely available speech model on rented machines in the EU. For each sentence it listens to a few seconds of your recording as a reference and imitates the sound. It learns nothing permanently in the process, and nothing flows to the maker of the model. Later we may adapt a copy of the model of our own with your recordings. That copy, too, stays with us and is deleted when the contract ends.

Themes

Social drama, intrigue, crime, thriller, fantasy, science fiction, adventure, quarrels, grief and implied violence. What is told is fiction for adults.

Love scenes

The Paris world contains love scenes. Players can choose a mode in which intimacy is told sensually and in concrete physical terms, without fading out. VALA does not use anatomical or pornographic language in doing so.

If you would like to exclude love scenes in the sensual mode, or an entire world, for your voice, please say so in your application. We are setting VALA up so that entire worlds and the sensual mode can be blocked for individual voices.

We cannot reliably exclude individual themes within a world, because the texts are created live. Before the contract we will tell you bindingly what we can promise.

Never

We do not use the voice for political advertising, for third-party advertising or for news. Nor does it imitate living persons. Within the game itself there may be invented politics and historical figures, a Roman senator for example.

Rights: a buyout for VALA only

For a fixed sum we buy the right to build an AI voice from your recordings and to use it for VALA only. We do not agree an ongoing revenue share; statutory remuneration claims remain unaffected.

This covers

  • the recordings as a voice reference and for adapting a speech model of our own, only within our closed system
  • the speech generated with them, in all present and future VALA worlds
  • all devices on which VALA runs (box, app, computer), with no territorial restriction
  • advertising for VALA, such as trailers, audio samples and adverts

We acquire no exclusivity over your voice. You go on working for whomever you wish, with AI too. Only the recordings you make for us are exclusive: you do not pass these files on to others.

Two prices

Please quote two prices for the right of use. On the basis of your quote we decide which option we choose, and we tell you before the contract. A narrator’s voice cannot be replaced overnight. That is why Option B provides for a renewal and a transition:

Option A, term: unlimited in time

After that: No further contractual payment.

Option B, term: 5 years

After that: The 5 years run from the first public use, and at the latest from six months after delivery. VALA may renew for 5 years at a time at a price agreed in advance. Without renewal, a 12-month transition follows; it is included in the price. After that we delete the recordings and the AI voice within 30 days.

What we promise

  • No passing on

    We neither sell nor lend your recordings or the AI voice to others. Technical service providers, such as the data centre, work solely on our behalf. We name one exception openly: if VALA is sold as a whole, the contract and the rights go with it. All limits and commitments then continue to apply unchanged.

  • No third-party training

    Your recordings do not train any third-party or public AI models.

  • No blending

    We do not blend your voice with other voices.

  • Language

    Your AI voice speaks German and, later, other languages as well.

  • Servers in the EU

    The recordings and the AI voice are kept only on servers in the EU that we operate or rent ourselves, currently in Germany or Finland.

  • Approval

    You hear your AI voice before it is released and can ask for improvements.

  • Labelling

    Players are told that the voices are AI-generated, as the EU AI Act requires.

  • Credit

    With your name in the credits and on our website, or without, as you prefer. We write “AI voice based on the voice of …”. That way nobody believes you spoke every sentence yourself. We advertise with your name only if you agree.

  • Liability

    If claims are brought against you over content that VALA generates with your AI voice, we indemnify you.

  • Misuse

    If we learn of any misuse of your voice, we inform you without delay.

Payment

  • We pay the recording fee 14 days after complete and defect-free delivery.
  • We pay the fee for the right of use as soon as your AI voice can be heard in public. We pay it six months after delivery at the latest, even if VALA has not yet been released by then.
  • We pay the artists’ social security contribution (Künstlersozialabgabe) on top. It does not reduce your fee.

If it does not work out

  • If we do not succeed in making a good AI voice from your recordings, either side may end the contract. You then keep the recording fee. The fee for the right of use lapses, and we delete everything.
  • Should we one day discontinue VALA, we delete the recordings and the AI voice. Fees already paid remain yours.

How to apply

Please send:

  1. the role, or the two roles, you are applying for
  2. an audio sample of the sample text below, as WAV or MP3, unprocessed and without music, in the recording set-up you would also use for the job
  3. your quote in two parts, net:
    • Recording: about 60 minutes from a script, including studio, editing into individual files and one round of pick-ups
    • Right of use: one price each for Option A and Option B
  4. if you wish: content you would like to exclude for your voice

We use your audio sample for selection only. We do not pass it on. In the second round we ask individual applicants whether we may test their sample once in our system. This happens only with your consent in text form, for example by email; without it, no sample enters our AI system. If no contract is concluded, we delete the sample and the test data no later than 30 days after the selection decision. How we process your application and audio sample is explained in our privacy policy, section 5.

Sample text

The sample text takes about two minutes. Please record all parts in one file, with a short pause between them. Do not speak the emotion cues before the colon. If you are applying for two roles, please send one sample per role. The text itself is in German, because that is what will be recorded.

Sample text

about 2 minutes

Part 1 · Narration, calm

Der Regen hat aufgehört, als die Kutsche vor dem Hôtel Beaumont hält. Elise de Beaumont, Comtesse de Valençay, steigt aus und zieht das schwarze Spitzentuch enger um die Schultern. Auf der Rue Saint-Honoré rufen die Händler die letzten Preise des Tages, und vom Palais-Royal klingt eine Geige herüber. Im Salon wartet Madame Émilie Pinchard, die Hände im Schoß gefaltet, als hätte sie seit Stunden nicht gewagt, sich zu rühren.

Part 2 · Character speech with emotion; please make each clearly distinct

  • happy:«Der Comte de Valmont hat zugesagt! Er kommt heute Abend, und er bringt die Herzogin d’Orsigny mit!»
  • sad:«Victor hat mir nie gesagt, wohin er an diesem Morgen ritt. Und ich habe ihn nicht gefragt.»
  • angry:«Sie wagen es, in meinem eigenen Salon über meinen Mann zu sprechen? Verlassen Sie sofort dieses Haus!»
  • tense, whispered:«Leise … da ist jemand auf der Treppe. Löschen Sie die Kerze, schnell.»
  • tender:«Bleiben Sie noch einen Augenblick. Nur einen. Ich habe so lange darauf gewartet.»

Part 3 · Names and numbers

Capitaine Henri Brossard, Doktor Augustin Boirel, Geneviève d’Orsigny und Lady Margaret Thornton treffen sich im Faubourg Saint-Germain. Am 14. März 1826, um halb elf, zahlt Monsieur Lefèvre 3.750 Francs. In der Subura des alten Rom sucht Marcus derweil nach Tertia, der Kräuterfrau.

Part 4 · Other worlds, calmly narrated

Die Schleuse öffnet sich mit einem leisen Zischen. Hinter dem Glas dreht sich langsam ein blauer Planet. Weit davon entfernt, am Rand des Nebelwalds, hebt die alte Hüterin ihre Laterne und lauscht.

VALA Worlds · Call for voices 2026

Process and dates

We accept applications until

7 December 2026

Apply by email
  1. Selection

    We listen to all samples and, by 18 December 2026, invite a shortlist to a video call. Everyone else will be told by then that they have not been selected.

  2. Test

    With your consent in text form (for example by email), we test your audio sample once in our system.

  3. Contract

    In writing. It puts the points of this call into effect; the signed contract is binding. You receive the draft at least 14 days before signing, so that you can have it reviewed at your leisure.

  4. Recording

    From our script, self-directed. Delivery within 4 weeks of your having the script and the contract.

  5. Approval

    You hear your AI voice before it is released.

Client and contact

Your contact is Stefanie Jasser, founder and managing director.

kontakt@valaworlds.de

I am happy to answer questions by email or in a short video call, before you apply as well.

VALA Worlds UG (haftungsbeschränkt) · Küstriner Str. 7a, 25421 Pinneberg, Germany · HRB 20137 PI, Amtsgericht Pinneberg (Pinneberg Local Court)