Tools / AI Avatar & Talking Head
D-ID
Talking-head avatars from a photo, plus API for personalised video
Disclosure: the Visit Website button may be an affiliate link — it supports the directory at no extra cost to you. Saving is private: it is stored only in your browser.
Facts checked · Nov 2026 — the pricing, free-plan and watermark facts on this page were checked against D-ID's official pages. See how we verify and the listing changelog.
Price alerts for D-ID
AI video tools change prices and credit rules without warning — Kling has raised prices roughly 3× before. Get emailed when D-ID's price, credits or free plan changes.
Our verdict
Best for turning a single photo into a talking presenter at the lowest entry price in the avatar category — and for developers who want that same trick inside their own product, via its API.
D-ID at a glance
| Starting price | Paid plans from around $5–6/month for entry tiers |
| Pricing type | Freemium |
| Free plan | Yes |
| Free export allowed | ✓ |
| No watermark on free export | ✕ |
| API available | ✓ |
| Category | AI Avatar & Talking Head |
| Entry price estimate (our calculator) | $5/month, subscription-based |
| How the billing works | Entry Lite tier; higher tiers add minutes and remove branding |
What D-ID really costs
D-ID bills in minutes, with one sharp edge: every video's length is rounded up to the nearest 15-second interval before deduction, per its pricing page. A 61-second video costs a minute and a quarter. Short, punchy presenter clips are therefore the efficient shape for this tool, and a script that drifts 10 seconds over a quarter boundary quietly costs you 25% more than its runtime.
The second rule is expiry: minutes do not accumulate. They renew each month and unused minutes become void, and API generations draw from the same single balance as studio work. Because each plan tier is sold in several minute packages, there is no single per-minute price to derive honest cost cards from — the rate depends on the package you chose. Budget by picking the package that matches your monthly finished minutes, including the rounding, rather than by headline price alone.
No cost cards are shown because D-ID sells each tier in multiple minute packages at different effective rates, and no single per-minute rate is published on its official pricing page, checked Nov 2026. The rounding and expiry rules above are from that page.
What is D-ID?
D-ID animates a still photo or avatar into a speaking presenter. Upload one image, type or record a script, choose a voice and language, and the face delivers it. That photo-to-talking-video trick is what made D-ID widely known, and it remains the cheapest credible way into presenter video: its own site prices the Lite plan from $4.70 a month billed annually, far below the business avatar platforms.
The second half of D-ID is the API. Businesses use it to generate personalised talking videos at scale — a presenter who greets each customer by name, for example — and D-ID's pricing page confirms API minutes draw from the same balance as the studio. The trade-off for the price is polish: our listing notes that avatar realism trails premium rivals on close inspection, and the studio is simpler than HeyGen's or Synthesia's. For a photo that needs to talk, cheaply and at volume, that is usually the right trade.
How D-ID works
- 1Start the trial — D-ID's official site describes a 14-day trial with 3 minutes of usage, and trial videos carry a full-screen watermark.
- 2Upload a photo or pick an avatar. A clear, front-facing photo gives the cleanest talking head.
- 3Write the script and choose a voice and language, or upload your own audio.
- 4Generate the video. Its length is deducted from your minutes, rounded up to the nearest 15 seconds — a 70-second video bills as 75 seconds.
- 5Review the face closely before publishing: mouth movement and realism are what separate a keeper from a re-do.
- 6Export, or move to the API with the key from your account page when you need the same video generated for many recipients.
Core features
Photo-to-talking-video
One still image becomes a speaking presenter. No video shoot, no green screen — the lowest-friction route to a talking head in this category, and the feature D-ID is known for.
Avatar studio
A browser studio for individual creators to build presenter videos with scripts, voices and languages. It is deliberately simpler than the big training platforms, which is a virtue at this price.
Developer API
A documented API for generating videos programmatically, including personalised videos at scale. D-ID's pricing page notes that API usage deducts from the same minutes balance as the web studio, so there is one pool to budget.
Multiple voices and languages
A library of voices across many languages for narration, so the same presenter can front localised versions of a video without re-shooting anything.
Video translation
D-ID also translates existing videos, keeping the speaker recognisable while the voice and lip movement adapt to the new language — covered on D-ID's own site as a separate workflow from the studio.
Interactive agent-style avatars
Beyond pre-rendered video, D-ID offers real-time conversational avatar products. That is a different buying decision from studio videos, but it shares the same account and API family.
The free plan — the full truth
D-ID's free route is a trial, and it is honest about what a trial is for. D-ID's official site describes 14 days with 3 minutes of usage. Our verified data adds the detail that matters: trial and free output is watermarked — D-ID's pricing page goes further and says a full-screen watermark appears for trial users. Three minutes is enough to animate your photo, hear your script delivered, and decide whether the realism passes for your audience. It is not a publishing budget.
What the trial does let you do is export — our truth data verifies free export is allowed, so you can download the watermarked result and show it to a team. The upgrade logic is then simple: Lite, from $4.70 a month billed annually, is among the cheapest entries in this whole catalogue, and D-ID's own site notes the commercial-use licence starts at the Pro plan, with Lite output still carrying the watermark. Know which of those two facts — price or licence — is driving your purchase before you pick a tier.
- • Trial: 14 days with 3 minutes of usage, per D-ID's official site
- • Trial users get a full-screen watermark, per D-ID's pricing page
- • Trial and Lite output carries a D-ID watermark (verified)
- • Video length is rounded up to the nearest 15 seconds against your minutes
- • Unused minutes do not accumulate — they renew monthly and then are void, per the pricing page
D-ID pricing plans
| Plan | Price | Credits included | Best for |
|---|---|---|---|
| Trial | $0 (14 days) | 3 minutes of usage, per D-ID's official site; full-screen watermark on output | Animating your own photo once and judging whether the realism passes |
| Lite | $4.70/month billed annually ($5.90 monthly) | Minute package chosen at checkout; minutes deducted in 15-second increments; output watermarked | The cheapest entry to presenter video for personal and test projects |
| Pro | $16/month billed annually | Larger minute packages; premium presenters at 1080p per D-ID's FAQ; commercial-use licence starts here, per D-ID's official site | Publishers and businesses who need commercial rights and better presenters |
Prices are from D-ID's official site and pricing pages, checked Nov 2026. Each self-serve tier is sold in several minute packages, so the minute allowance for your exact package is set at checkout; the pricing page's own FAQ explains the package structure.
Pros
- + The lowest entry price in the avatar category by a wide margin
- + A single photo becomes a presenter — no filming at all
- + Strong, documented API with one shared minutes balance
- + Trial is long enough and honest enough to judge realism properly
- + Translation and interactive-agent products sit in the same account family
Cons
- − Trial and Lite output is watermarked — clean output needs a higher tier
- − Commercial-use licence starts at Pro, per D-ID's own site — Lite is not the business tier its price suggests
- − Avatar realism trails premium rivals on close inspection
- − Minutes expire monthly and lengths round up to 15-second blocks
- − Studio features are simpler than HeyGen or Synthesia
Best for
- ✓ Creators who need a talking photo on a very small budget
- ✓ Developers generating personalised presenter videos through an API
- ✓ Businesses prototyping avatar video before committing to a premium platform
- ✓ Anyone localising a presenter's message across voices and languages
Not ideal for
- ✕ Brands for whom presenter realism is the whole point
- ✕ Users who need a full training platform with courses and SCORM
- ✕ Anyone expecting commercial rights at the Lite price
D-ID alternatives
See all alternatives →HeyGen is the first alternative in the avatar category of our catalogue: a more polished presenter studio with stronger realism, better when the avatar fronts your brand publicly — at several times D-ID's entry price. Synthesia is the pick for corporate training libraries, with the deepest learning ecosystem. Colossyan also aims at training, with a free Starter tier and strong localisation. Choose D-ID when price or the API is the deciding factor; choose the others when the presenter's realism or the training workflow is. Their pages below carry the same verified treatment.
HeyGen
HeyGen turns a script into a video of a realistic digital presenter, with a large avatar library, voice cloning and translation features. It is widely used for marketing videos, product explainers and social content where a human-like presenter is needed without filming.
Synthesia
Synthesia is an enterprise-focused avatar video platform used heavily for training and internal communication. It offers a large stock avatar library, personal avatars, multi-language narration and compliance and admin features that larger organisations require.
Colossyan creates presenter-led videos aimed at workplace learning, with scenario-style content, many languages and features for localising training across regions. It competes most directly with Synthesia in the corporate training niche.
Frequently Asked Questions
Is D-ID free?
There is a 14-day trial with 3 minutes of usage, per D-ID's official site. Trial output carries a full-screen watermark. Ongoing use needs a paid plan, from $4.70/month billed annually for Lite.
Does D-ID watermark free videos?
Yes — verified in our data for trial and free output, and D-ID's pricing page states a full-screen watermark appears for trial users. Lite output is also watermarked; the watermark is one of the things higher tiers remove.
Can I use D-ID videos commercially?
According to D-ID's official site, the commercial-use licence starts on the Pro plan ($16/month billed annually). Trial and Lite are not the tiers to publish paid client or product work from — check D-ID's current terms for your exact use.
Does D-ID have an API?
Yes — verified in our data. D-ID's pricing page confirms API minutes are deducted from the same balance as the web studio, and its documentation lives in its developer hub. It is one of the main reasons businesses choose D-ID.
How does D-ID count minutes?
By finished video length, rounded up to the nearest 15 seconds: a 1-minute-10-second video consumes 1 minute 15 seconds, per the pricing page. Minutes renew monthly and unused minutes become void.
D-ID or HeyGen?
D-ID wins on entry price and on turning a single photo into a speaker, with a strong API. HeyGen wins on presenter realism and studio polish for public-facing brand video. If the video is personalised at scale or budget-led, D-ID; if it is your homepage presenter, HeyGen.