Tool review
ElevenLabs review: narration and paid image-to-video in one workspace
ElevenLabs offers narration, Studio editing and image-to-video through partner models. Free audio credits do not include video generation.
Best for
Narration on a memorial, a birthday film, a product video or a short social clip, when a good voice is the difference between watched and skipped.
Free plan with 10,000 credits a month and no commercial licence. Starter is $6 a month, with a commercial licence. Check current pricing.
What users say
Third-party ratings
Each line is someone else’s figure, copied from the linked page on the date shown. pics2vids does not publish its own score.
Often praised The linked Trustpilot profile includes positive comments about voice quality and ease of use.
Often complained about Negative comments on the same profile include billing and cancellation issues. Self-selected reviews do not establish how common these problems are.
Plan the voice track and the generated shots against the same credit balance.
Source photo, then the clip
Pictures for this page
A sample prompt
Three sentences of warm, slow narration over a photograph of a family member. One voice. No music.
I have not used ElevenLabs for a private family tribute. I did make a short fictional village film for this page; it is in the demo section below. Prices and credit limits are from the public pricing and docs pages, checked on the date below.
What it is good at
- The free plan is a real $0 tier, with 10,000 credits a month, usable for a short sample.
- A paid Starter plan adds a commercial licence and instant voice cloning at a low monthly price.
- Studio can export MP3 or WAV, which drops straight into CapCut or any timeline.
- Language support depends on the speech model; listen to the selected voice before using it.
Where it falls short
- Video generation uses paid credits shared with other features; check the model cost first.
- The free plan has no commercial licence and no voice cloning.
- Speech credit usage varies by model and voice; a fixed characters-to-minutes conversion is only an estimate.
- Read the dated third-party rating summary below and check the written reviews before paying.
ElevenLabs combines AI speech with image and video generation. For a photo-led film, you can use it for narration or for a generated shot from a reference photograph. Those features use the same subscription credit balance.
Public pages were checked on 14 November 2026.
What it does
The Image & Video documentation describes uploading reference images, choosing a partner model and generating a clip. Settings such as duration and resolution affect credit use. Video generation needs a paid plan. Free users can generate images, limited to three image requests a day.
The pricing page lists Text to Speech, Speech to Text, Sound Effects, Voice Design, Music, Studio and Dubbing. The one most useful here is Text to Speech, with Studio as the workspace that holds a script and lets you assign voices to different lines.
Studio is described in the ElevenCreative docs as “a workspace for creating audio and video projects” with a multi-track timeline. You can import a script from EPUB, PDF, DOCX, TXT, HTML or a URL, assign voices to paragraphs, and export MP3, WAV or a video with captions burned in. For this site, the export you want is MP3 or WAV, dropped into CapCut or another editor over your own picture track.
What the free plan really is
The Free plan includes 10,000 credits a month and three Studio projects. Check current Studio export quality and watermark restrictions before publishing.
The pricing page estimates roughly ten minutes of text-to-speech on the free allowance. Usage varies by speech model and voice, and video generation has separate model rates. Check the cost shown before generating.
The free plan does not include a commercial licence, instant voice cloning or dubbing. A commercial use on the free plan is not allowed by the terms. Read the licence before you post a paid client video made on free credits.
Paid plans, as listed
On the same date the pricing page showed, in US dollars:
| Plan | Monthly | Credits |
|---|---|---|
| Starter | $6 | 30,000 |
| Creator | $22 | 121,000 |
| Pro | $99 | 600,000 |
| Scale | $299 | 1.8 million |
| Business | $990 | 6 million |
Introductory discounts are not the ongoing subscription price. Starter adds a commercial licence, instant voice cloning, 20 Studio projects, music commercial use and dubbing-studio access. Creator adds professional voice cloning, which takes a longer sample for a closer match. Pro lists 44.1 kHz PCM output via API and 192 kbps audio as separate features. The two top rows are team plans, with workspace seats and professional voice clones for staff voices.
Prices are in US dollars. The public pricing page I checked did not list a GBP figure. A UK checkout can add tax. Read the price in your own account.
Pairing it with the rest of the site
A memorial film built in CapCut with a stillness-first timeline can use a narration track if the family wants one. The route is:
- Build the timeline in CapCut, with each photograph held for the number of seconds that feels right.
- Write the narration as plain text.
- In ElevenLabs Studio, paste the text, pick a voice, and export an MP3.
- Drop the MP3 into CapCut on an audio track. Adjust the picture timing to the voice.
The same pattern works for a birthday film, a short product push, or a 9:16 social clip. For memorial use specifically, read the long guide at how to make a memorial tribute.
An explainer made in ElevenLabs Studio
I made this 27-second explainer in ElevenLabs Studio from a short script. The six video clips were generated with the Gemini Omni Flash 1.2 model inside ElevenLabs Image & Video, the narration is the James voice, and the music is Eleven Music. I then trimmed and captioned it myself.
AI clips can still show odd details, such as a slightly blurred hand and made-up handwriting in the notebook shot.
The video clips cost the most: about 2,140 credits for each 5-second 720p clip on 14 November 2026. The narration and the music used only a few hundred credits each. Against the 90,000 credits available on my Starter plan that month, making this took roughly a fifth.
Demo: A year in Little Haven
I made this short film to show what ElevenLabs can add to a photo-led timeline. The narration is the Voice Library voice James, generated with Eleven v4. The instrumental bed came from Eleven Music, and the village atmospheres came from Sound Effects v3. The stills were generated with ChatGPT. I cut the film with ffmpeg.
This demo used about 2,700 credits on 14 November 2026, on a Starter plan. That is what we used for this film and the samples below, not a price you should plan around. Rates vary by model and can change.
Eleven v4 ignored the speed setting, so I lengthened the pauses between sentences afterwards. The words themselves were not stretched.
Side-by-side samples
Each spoken sample is an AI-generated voice.
The same closing line, in three Voice Library voices, all on Eleven v4, with no audio tags:
It wasn’t a year of big news. Just the ordinary days we didn’t want to lose.
An expressive line with audio tags, again James on Eleven v4. The tags were in the input; they are not part of the line itself.
The cake never quite survived the afternoon. It never does. And somehow, that’s the point.
The same closing line in Bulgarian, generated as text-to-speech with James and Eleven v4, with the language set to Bulgarian. This is the TTS fallback, not a Dubbing v2 pass, so it does not keep the English timing of the original take.
Не беше година на големи новини. Просто обикновените дни, които не искахме да изгубим.
A sound-effects sample from SFX v3, requested without intelligible speech. The input was:
Village summer fete from a little way off: a small brass band, light crowd murmur, bunting fluttering in a breeze, a dog barking once in the distance, no intelligible speech.
A note on voice cloning
ElevenLabs lets you clone a voice, instant on Starter and professional on Creator. For a tribute, the question is permission, not capability. Cloning the voice of a person who has died, from a recording, is legally and ethically different from cloning your own voice to narrate a long project you own. Read ElevenLabs’ terms before you upload a sample that is not yours, and confirm the speaker has given the permission required by the service; family approval alone may not satisfy its rules. A plain, well-read ElevenLabs voice over a photograph is kinder and safer than a copy of a voice no one agreed to recreate.
Sources
Checked 14 November 2026:
Questions people ask
Can ElevenLabs animate a photograph?
Yes. The Image & Video workspace can generate video from reference images using partner models. Video generation requires a paid plan, while free users have limited image generation. Check the model, duration and credit cost before generating.
Is ElevenLabs free?
The free plan is listed at $0 with 10,000 credits a month. It does not include a commercial licence or voice cloning. The pricing page estimates about ten minutes of text-to-speech on Free; actual usage depends on the model and voice.
Can I clone my own voice?
Yes, on the Starter plan and above, listed at $6 a month on monthly billing. Starter includes instant voice cloning. Professional voice cloning, which uses a longer sample for a better match, needs Creator at $22 a month, with annual options subject to checkout terms.