Démarrage rapide
Ce guide couvre toutes les fonctionnalités de D-ID :
- Commencer — Create your account and set up the Studio
- Comment utiliser le traducteur vidéo D-ID — Turn one video into dozens of languages
- Comment utiliser l'API D-ID Talking Head — Generate talking head videos from code
- Comment utiliser les intégrations D-ID — Connect D-ID to the tools you already run
- Comment utiliser les campagnes vidéo D-ID — Send personalized video to hundreds of people at once
- Comment utiliser les agents d'IA D-ID — Build an assistant that answers from your own documents
- Comment utiliser les avatars générés par l'IA de D-ID — Create a presenter from a written description
- Comment utiliser la conversion photo-vidéo D-ID — Make any single photo speak your script
- How to Use D-ID Voice Cloning — Give your avatar your own voice
- How to Use D-ID Emotion and Expression Control — Set the mood of every line your avatar delivers
Temps nécessaire : 5 minutes par fonctionnalité
Vous trouverez également dans ce guide : Conseils de pro | Erreurs courantes | Dépannage | Tarification | Alternatives
Pourquoi faire confiance à ce guide ?
I have used D-ID for eight months and tested every feature covered here.
Ce tutoriel est issu d'une véritable expérience pratique, et non de captures d'écran fournies par le vendeur.
D-ID is one of the most capable AI video générateurs disponible dès aujourd'hui.
Most people only ever touch the photo upload button.
Ce guide vous montre comment utiliser chaque fonctionnalité principale.
Étape par étape, avec captures d'écran et conseils de pro.
Tutoriel D-ID
This How to Use D-ID tutorial walks through every feature, from your first login to the D-ID API.
A FAIT
Turn one photo into a talking avatar in under a minute. D-ID covers 119 languages, clonage vocal and a real time API for digital humans. Start on the free Trial plan — no credit card needed.
Premiers pas avec D-ID
Avant d'utiliser une quelconque fonctionnalité, veuillez terminer cette configuration unique.
Cela prend environ trois minutes.
Étape 1 : Créez votre compte
Go to the D-ID website and click “Get Started”.
Sign up with Google, Microsoft, or an email and password.
✓ Point de contrôle: Vérifiez votre boîte de réception pour l'e-mail de confirmation.
Step 2: Open Creative Reality Studio
D-ID runs in the browser, so there is nothing to install.
Log in and the Studio loads with your video library on the left.
Here is what my workspace looks like after a few months of use:
✓ Point de contrôle: Vous devriez voir le tableau de bord principal.
Step 3: Set Your Defaults
Pick a canvas layout for mobile, social media, or présentations.
Set your default language and voice provider before you start writing.
Trial credits appear in the top right corner of the page.
✅ Terminé : Vous êtes prêt à utiliser n'importe quelle fonctionnalité ci-dessous.
Comment utiliser le traducteur vidéo D-ID
Traducteur vidéo lets you publish one recording in dozens of languages without reshooting it.
Voici comment l'utiliser étape par étape.
Étape 1 : Téléchargez votre vidéo source
Open the Video Translator tab and drop in your finished clip.
Files under five minutes process fastest.
Étape 2 : Sélectionner les langues cibles
Choose from the 119 languages D-ID supports, with accents included.
You can add several at once instead of running separate jobs.
Voici à quoi cela ressemble :
✓ Point de contrôle: Each language appears as a separate item in your library.
Étape 3 : Générer et télécharger
Click Translate and wait for each version to render.
Every file comes back in MP4 format, ready to upload.
✅ Résultat : One video is now ready for every market you sell into.
💡 Conseil de pro : Translate the shortest cut of your video first, then check lip sync before running the long version.
Comment utiliser l'API D-ID Talking Head
API Talking Head lets you generate talking head videos straight from your own code.
Voici comment l'utiliser étape par étape.
Étape 1 : Générer une clé API
Open your account settings and create an API key.
Store it as an environment variable, never inside client code.
Step 2: Build the Request Payload
Your payload needs a source image URL and a script object.
The script accepts plain text or an audio file you host.
Voici à quoi cela ressemble :
✓ Point de contrôle: The API returns a 201 response with a talk ID.
Step 3: Send the Call and Poll
POST the payload, then poll the returned ID until status is done.
The API can handle tens of thousands of requests in parallel.
✅ Résultat : You can generate digital humans at scale from your own app.
💡 Conseil de pro : Cache the source image URL on your side — repeated uploads are the slowest part of video production.
Comment utiliser les intégrations D-ID
Intégrations D-ID lets you connect D-ID to the tools your team already opens daily.
Voici comment l'utiliser étape par étape.
Étape 1 : Ouvrez la page des intégrations
Find Integrations in the left sidebar of the Studio.
Chat, storage, and automation services each get their own card.
Step 2: Pick Your Provider
Select the service you want to connect and click Add.
Chat providers let you attach an avatar to an existing bot.
Voici à quoi cela ressemble :
✓ Point de contrôle: The provider card moves to your Connected list.
Step 3: Authorize and Test
Approve the permission screen, then run the built-in test call.
A green status badge means the connection is live.
✅ Résultat : Your avatar now works inside tools your team already uses.
💡 Conseil de pro : Connect one provider at a time and test it, or you will not know which integration broke.
Comment utiliser les campagnes vidéo D-ID
Campagnes vidéo lets you send a personalized video to every name on a list.
Voici comment l'utiliser étape par étape.
Étape 1 : Importez votre liste de contacts
Import a CSV with one row per recipient and a name column.
Clean the list first — bad names show up on screen.
Step 2: Add Variables to the Script
Write one script and wrap merge fields in curly braces.
D-ID swaps each field per recipient at render time.
Voici à quoi cela ressemble :
✓ Point de contrôle: The preview shows the first contact’s name in place.
Étape 3 : Générer et partager
Click Generate to build every version in one batch.
You get a landing page link for each contact.
✅ Résultat : Hundreds of personalized videos went out from one script.
💡 Conseil de pro : Keep the personalized part in the first six seconds, where people actually still watch.
Comment utiliser les agents d'IA D-ID
Agents d'IA D-ID lets you build assistants that answer questions from documents you provide.
Voici comment l'utiliser étape par étape.
Step 1: Select a Role and Write Instructions
Creating an agent starts with choosing a role and giving instructions.
Agents can simulate real people, fictional characters, or brand influencers.
Étape 2 : Télécharger les documents de connaissances
You can upload up to 5 documents as the knowledge base.
PDF, TXT and PPTX files work, and website URLs are supported.
Voici à quoi cela ressemble :
✓ Point de contrôle: Your documents show a green Processed label.
Step 3: Share or Embed
Share your agent with a link or embed it on your website.
Visitors can type a question or use voice input instead.
✅ Résultat : You have a live agent answering from your own knowledge.
💡 Conseil de pro : Upload one clean FAQ document instead of five overlapping ones — contradictions confuse agents fast.
Comment utiliser les avatars générés par l'IA de D-ID
Avatars générés par l'IA lets you create a presenter from a written description in seconds.
Voici comment l'utiliser étape par étape.
Étape 1 : Ouvrez le générateur d’avatars
Go to Avatars, then switch to the Generate tab.
This runs on text-to-image, so no photo is needed.
Step 2: Describe and Click Enter
Type a short description of age, clothing, and background.
Click Enter and D-ID’s AI returns four options to compare.
Voici à quoi cela ressemble :
✓ Point de contrôle: Four avatar variations appear side by side.
Étape 3 : Enregistrer dans votre bibliothèque
Pick the strongest result and save it.
Saved avatars stay available for every future video.
✅ Résultat : You own a reusable presenter that no competitor has.
💡 Conseil de pro : Describe the background separately from the person, or D-ID’s AI blends the two into mush.
How to Use D-ID Photo-to-Video Converter
Conversion photo en vidéo lets you turn one still photo into a talking avatar.
Voici comment l'utiliser étape par étape.
Étape 1 : Téléchargez votre photo
A single image is all it takes to create a speaking digital human.
Use a front-facing shot with both eyes clearly visible.
Étape 2 : Ajoutez votre script ou votre audio
Paste text and select a voice, or upload an audio file.
MP3, WAV, M4A, FLAC and MP4 audio all work.
Voici à quoi cela ressemble :
✓ Point de contrôle: The photo preview shows a play button over it.
Étape 3 : Générer la vidéo
Click Generate and the render usually finishes in under a minute.
Maximum length is five minutes per video.
✅ Résultat : A single photo now speaks your script on camera.
💡 Conseil de pro : Crop the photo so the head fills roughly a third of the frame before you upload it.
How to Use D-ID Voice Cloning
Clonage vocal lets you put your own voice behind any avatar you choose.
Voici comment l'utiliser étape par étape.
Étape 1 : Ouvrez la bibliothèque vocale
Select your avatar, then open the voice dropdown in the script box.
Click “Clone voice” at the bottom of the list.
Step 2: Upload or Record
Upload a clean audio file or record directly in the browser.
Thirty seconds of clear speech beats five noisy minutes.
✓ Point de contrôle: Your cloned voice reads the preview line back to you.
Step 3: Assign the Cloned Voice
Name your Instant Cloned Voice and save it.
It then appears as a voice option for every avatar.
✅ Résultat : Every future video can sound like you without recording.
💡 Conseil de pro : Record your sample in the same tone you use on camera, not your reading-aloud voice.
How to Use D-ID Emotion and Expression Controller
Emotion and Expression Control lets you control how each line of your script lands.
Voici comment l'utiliser étape par étape.
Étape 1 : Choisissez votre avatar
Open the video you are editing and click the avatar layer.
Expression settings sit under the script box.
Step 2: Choose an Expression
Pick happy, serious, surprised, or neutral for the delivery.
Match the expression to the line, not the whole script.
Voici à quoi cela ressemble :
✓ Point de contrôle: The preview thumbnail updates with the new expression.
Step 3: Add Pauses Where They Matter
Click the stopwatch icon in the text box to add a pause.
A half-second break before a key claim reads as confidence.
✅ Résultat : Your script now reads like a person, not a robot.
💡 Conseil de pro : Use neutral as your base and add one happy or serious moment, not six.
Conseils et astuces de pro pour D-ID
After eight months in the Studio, these are the habits that saved me the most time.
Raccourcis clavier
| Action | Raccourci |
|---|---|
| Annuler la dernière modification | Ctrl/Cmd + Z |
| Rétablir la dernière modification | Ctrl/Cmd + Maj + Z |
| Enregistrer le brouillon | Ctrl / Cmd + S |
| Duplicate selected layer | Ctrl / Cmd + D |
Fonctionnalités cachées que la plupart des gens ignorent
- Pause control: Click the stopwatch icon in the text box to add a pause mid-script.
- Canvas layouts: Switch format for mobile, social media, or presentations before you generate.
- Accents and styles: The 119 languages each carry accent and speaking style options.
- Visual layers: Add backgrounds, videos, and text on top of any avatar.
- Parallel API calls: The D-ID API handles tens of thousands of requests at once.
Here are the benefits I lean on most:
Erreurs courantes à éviter en matière de D-ID
Mistake #1: Feeding it a bad photo
❌ Faux : Uploading a blurry side-angle selfie and blaming the output.
✅ À droite : Use a sharp, front-facing photo with both eyes visible.
Mistake #2: Writing a script that runs too long
❌ Faux : Pasting a ten minute script and waiting for an error.
✅ À droite : Keep each video under the five minute limit and split longer scripts.
Mistake #3: Exposing your API key
❌ Faux : Hardcoding the API key in front-end JavaScript.
✅ À droite : Call the D-ID API from your server and keep the key in environment variables.
Dépannage D-ID
Problem: The video is stuck on processing
Cause: Long scripts and busy render queues slow generation down.
Réparer: Refresh the page, then shorten the script and generate again.
Problem: The API returns a 401 error
Cause: Your API key is missing, expired, or wrongly encoded in the header.
Réparer: Generate a fresh key in settings and check the Authorization header format.
Problem: The lip sync looks off
Cause: Background noise or heavy compression in the uploaded audio file.
Réparer: Re-record in a quiet room and upload WAV instead of a low bitrate MP3.
📌 Note: If none of these fix your issue, contact D-ID support.
Qu'est-ce que le D-ID ?
A FAIT is an AI video generator that turns photos and text into talking avatars.
Its Creative Reality Studio is a self-service platform for creating videos with moving and talking avatars.
The Studio combines deep-learning face animation with generative AI text and text-to-image tools.
Think of it like a film studio where the actors are digital humans.
Regardez ce bref aperçu :
Il comprend les fonctionnalités clés suivantes :
- Traducteur vidéo : Translates a finished video and re-syncs the lip movement
- API Talking Head : Builds talking head videos from an image plus text or audio
- Intégrations D-ID : Pushes avatars into chat tools, LMS platforms and slide decks
- Campagnes vidéo : Spins one script into hundreds of personalized videos
- Agents d'IA D-ID : Autonome assistants that answer from knowledge you upload
- Avatars générés par l'IA : Creates brand-new presenters from a text prompt
- Conversion photo en vidéo : Turns one still photo into a speaking digital human
- Clonage vocal : Copies your voice from a short recording or audio file
- Contrôle des émotions et des expressions : Sets happy, serious, surprised or neutral delivery per line
Pour une analyse complète, consultez notre Examen D-ID.
Tarification D-ID
Here is what D-ID costs in 2026:
| Plan | Prix | Idéal pour |
|---|---|---|
| Procès | 0 $/mois | Testing the Studio before you pay |
| Lite | 4,7 $/mois | Solo creators making a few clips |
| Pro | 16 $/mois | Regular video production and voice cloning |
| Avancé | 108 $/mois | Teams running campaigns and API calls |
| Entreprise | Tarification personnalisée | High volume and custom agreements |
Essai gratuit : Yes — the Trial plan is $0/month with limited credits.
Garantie de remboursement : Not advertised, so test on Trial before you upgrade.
💰 Meilleur rapport qualité-prix : Pro at $16/month — it unlocks voice cloning without the Advanced jump.
D-ID vs Alternatives
How does D-ID compare? Here is the competitive landscape:
| Outil | Idéal pour | Prix d'entrée | Notation |
|---|---|---|---|
| A FAIT | Photo-to-video and API | 4,7 $/mois | ⭐ 4.2 |
| HeyGen | Avatar quality | ~$29/mo | ⭐ 4,6 |
| Synthesia | Équipes d'entreprise | ~$29/mo | ⭐ 4,6 |
| Colossian | Training video | ~$27/mo | ⭐ 4,5 |
| Veed | Editing plus avatars | ~$19/mo | ⭐ 4,5 |
| Elai | Slides to video | ~$23/mo | ⭐ 4,4 |
| Vidnoz | Free tier volume | ~$19/mo | ⭐ 4,3 |
| Deepbrain | Stock realism | ~$24/mo | ⭐ 4,4 |
Sélection rapide :
- Meilleur résultat global : HeyGen — the avatars still look the most natural.
- Meilleur budget : D-ID — paid plans start at $4.7/month.
- Idéal pour les débutants : Vidnoz — templates and a free tier to learn on.
- Idéal pour les développeurs : D-ID — the API is the deepest in this group.
🎯 Alternatives au D-ID
Looking for D-ID alternatives? Here are the options worth your time:
- 🚀 HeyGen : Fastest avatar library and the smoothest translation workflow for marketing teams.
- 🏢 Synthesia : Enterprise favourite with strict brand controls and a huge stock avatar set.
- 👶 Colossyan : Built around training video, with scenario branching that beginners pick up fast.
- 🎨 Veed : Full editor first, avatars second — best if you also cut and caption footage.
- 🔧 Elai : Slide-to-video conversion and a clean API for developers on smaller budgets.
- 💰 Vidnoz : Generous free tier and hundreds of templates for low-volume social clips.
- ⭐ Cerveau profond : Strong realism on stock presenters, popular for news-style and kiosk video.
- 🧠 Synthesys : Voice-led tool where the audio quality matters more than the avatar.
- 📊 Première heure : Scales product and course video from spreadsheets rather than one-off edits.
- ⚡ Virbo : Mobile-first app from Wondershare for quick vertical avatar clips.
- 💼 Vidyard : Sales video hosting with tracking, if you record yourself more than you generate.
- 🎯 Fliki : Text-to-video with a big voice library, cheaper for faceless content.
- 🔊 Discourir: Voice reading and cloning tool, not a full video production suite.
- 🔥 Dans la vidéo : Prompt-to-video editing with stock footage baked into every template.
- 🛒 Créativité : Ad-focused: paste a product URL and get avatar ad variations back.
- 📱 Sous-titres IA : Phone app for editing, dubbing and lip-syncing your own face on camera.
Pour la liste complète, consultez notre Alternatives au D-ID guide.
⚔️ Comparaison D-ID
Here is how D-ID stacks up against each competitor:
- D-ID contre HeyGen: HeyGen has more polished stock avatars. D-ID wins on API depth and price.
- D-ID contre Synthesia : Synthesia suits large enterprise rollouts. D-ID is far cheaper to start testing.
- D-ID contre Colossyan : Colossyan is better for structured training. D-ID handles real-time chat avatars better.
- D-ID contre Veed : Veed is the stronger editor. D-ID is the stronger digital humans engine.
- D-ID contre Elai : Elai converts slides faster. D-ID gives you better emotion and expression control.
- D-ID contre Vidnoz : Vidnoz has a bigger free tier. D-ID gives cleaner output and a real API.
- D-ID contre Deepbrain : Deepbrain edges ahead on stock realism. D-ID lets you use any photo you own.
- D-ID contre Synthesys : Synthesys focuses on voice. D-ID covers voice plus avatar plus live agents.
- D-ID contre la première heure : Hour One is built for bulk catalogues. D-ID is more flexible for one-off video.
- D-ID contre Virbo : Virbo is easier on mobile. D-ID scales better once volume grows.
- D-ID contre Vidyard : Vidyard hosts and tracks recordings. D-ID generates the footage in the first place.
- D-ID contre Fliki : Fliki is cheaper for faceless clips. D-ID is the pick when you need a face.
- D-ID contre Speechify : Speechify only produces audio. D-ID pairs that audio with a talking avatar.
- D-ID contre InVideo : InVideo leans on stock footage. D-ID leans on avatars and photo animation.
- D-ID contre Creatify : Creatify is tuned for paid ads. D-ID covers a much wider set of use cases.
- D-ID contre IA de légende : Captions AI edits your own footage. D-ID creates presenters from scratch.
Commencez à utiliser D-ID dès maintenant
Vous avez appris à utiliser toutes les principales fonctionnalités de D-ID :
- ✅ Traducteur vidéo
- ✅ API Talking Head
- ✅ Intégrations D-ID
- ✅ Campagnes vidéo
- ✅ Agents IA D-ID
- ✅ Avatars générés par l'IA
- ✅ Conversion photo en vidéo
- ✅ Clonage vocal
- ✅ Emotion and Expression Control
Étape suivante : Choisissez une fonctionnalité et essayez-la dès maintenant.
La plupart des gens commencent par la conversion de photos en vidéos.
Cela prend moins de cinq minutes.
Foire aux questions
Comment fonctionne le D-ID ?
You upload a photo or pick an avatar, then add a script or audio file. D-ID’s AI animates the face, syncs the lips and returns an MP4 video.
Puis-je utiliser D-ID gratuitement ?
Yes. The Trial plan costs $0/month and lets you generate a limited number of videos. Paid plans start at $4.7/month once you need more credits.
D-ID possède-t-il une API ?
Yes. The D-ID API creates digital humans at scale from one image and text or audio. It supports real time streaming and parallel requests.
Comment réaliser une vidéo D-ID ?
Open Creative Reality Studio, choose a canvas layout, pick an avatar, paste your script, select a voice, then click Generate. Videos export as MP4.
Puis-je utiliser ma propre voix dans D-ID ?
Yes. Upload an audio file or record directly in the Studio to create an Instant Cloned Voice, then assign that voice to any avatar.
English 
Spanish 
Chinese 
German 
French 
Arabic 
Russian 
Portuguese 
Japanese 
Korean 
Italian 
Turkish 
Dutch 
Polish 
Indonesian 
Thai 
Vietnamese 