D-ID

The talking-photo pioneer: turn a picture into a presenter, with V4 Expressive Avatars that change sentiment, real-time Visual AI Agents and video translation. A free mini-course.

What it nails

• Turns one photo into a talking presenter
• V4 avatars with sentiment control
• Real-time Visual AI Agents in up to 4K

Watch out for

• Minutes don't roll over; watermark below Enterprise

D-ID made its name turning a single photo into a talking video. Today its Creative Reality Studio lets you "transform text, audio, or still images into high-quality videos with lifelike digital avatars", and its newest V4 Expressive Avatars (February 2026) deliver lines with a chosen sentiment: "Messages sound calmer when reassurance is needed, more confident when authority matters." D-ID also builds Visual AI Agents, real-time digital humans connected to any AI model.

Start here: the basics

1. Start the free trial: 14 days, with a full-screen watermark.
2. Pick a presenter: a stock V4 avatar, an avatar generated from text, or your own photo.
3. Write the script (up to about 5 minutes, roughly 700 words) or upload audio, and pick a voice. The magic-wand button lets GPT help write it.
4. Choose a sentiment on V4 avatars: Calm, Confident, Amicable, Energetic, Friendly, Professional or Frustrated.

Photo tips for a great talking photo

• "Facing camera, medium shot"
• "Neutral expression, closed mouth"
• "Minimum head size 200×200 pixels" and "Good and consistent lighting"
• "No face occlusions (hats, sunglasses, masks, visors, large earrings)"

Hello! This is my first talking avatar video.

Example by D-ID · source ↗ · D-ID's own first script, voiced with the Microsoft voice en-US-JennyNeural.

Power moves

1. Build a Visual AI Agent

D-ID's agents "respond naturally, carry out tasks, and integrate securely", work with "any LLM" and answer from your own knowledge base. V4 Expressive Visual Agents (March 2026) talk in real time with "low latency (sub-0.5-second) conversational turns" in up to 4K. D-ID's prompt structure is Persona, Key Rules, Off-topic Handling, Limitations, Proactive Lead, Fallback; keep answers under 400 characters, with no emojis or stage directions.

You are Chloe, a friendly customer support specialist who helps users troubleshoot and set up our product.

Example by D-ID · source ↗ · From D-ID's official agent guide: start with Persona, then Key Rules.

If asked about unrelated topics (e.g., aliens), respond with humor and steer back: 'Aliens? Haven't met any—yet!'

Example by D-ID · source ↗

Q: How do I reset my password? A: Open Settings menu, select Account, then choose Reset Password.

Example by D-ID · source ↗ · D-ID recommends a question-and-answer format for the agent's knowledge.

I don't have exact details, but you can check here: [URL].

Example by D-ID · source ↗

2. Translate your videos

Video Translate "clones the speaker's voice" and "adapts lip movements to match the new language".

3. Personalised video campaigns

Insert dynamic fields like {{contact.firstname}} (HubSpot) or |FNAME| (Mailchimp), and every viewer hears their own name.

4. Agentic Videos

Launched in April 2026, Agentic Videos "combine the most powerful format for content delivery: video, with real-time, interactive AI", so viewers can ask the video questions.

Plans and minutes

D-ID's plans are Free Trial (14 days), Lite, Pro, Advanced and Enterprise, starting at $5.90 a month. From D-ID's own pricing FAQ and terms:

• "Minutes do not accumulate, they are renewed every month and unused minutes become void."
• Video length is rounded up to the next 15 seconds: "a 1-minute and 10-second video will consume 1 minute and 15 seconds."
• Watermarks: full-screen on the trial, a D-ID mark on Lite, a generic AI mark on Pro; only Enterprise removes it fully.
• Payments are "irrevocable and nonrefundable", and you must cancel 15 days before the billing period ends.

Make the pieces in the Lab

AI Whisperer Lab does not make talking avatars, but it runs the parts around them by the run, with credits that never expire:

• Voiceovers with ElevenLabs, the premium voice provider D-ID supports: text to speech, voice cloning and speech to speech in the Audio studio.
• Presenter portraits and scenes with Nano Banana Pro, GPT Image 2 or Seedream in the Image studio.
• Cinematic b-roll with Veo 3.1, Kling 3.0 or Seedance in the Video studio.

FAQ

Sources: D-ID Creative Reality Studio · V4 Expressive Avatars · Visual AI Agents · D-ID pricing FAQ · Agent prompt guide · Ethics