Depthify.ai
Depthify.ai turns 2D photos and videos into 3D spatial media for Apple Vision Pro and Meta Quest. It estimates depth, builds a stereo pair, then encodes MV-HEVC or HEIC, either in the cloud or entirely on your Mac.
What is Depthify.ai?
Depthify.ai is a conversion tool with one narrow purpose: taking flat 2D photographs and videos and turning them into 3D spatial media that plays back on an Apple Vision Pro or a Meta Quest. The site's own framing is to add depth to any photo or video for viewing on those two headsets.
The pipeline is described openly, in three stages. First, a monocular depth network predicts the metric depth of every pixel in every frame. Second, the RGB image and that depth map are combined into the left-eye and right-eye views of a stereo pair, by computing the estimated disparity between the eyes. Third, the result is encoded as an .HEIC image or an MV-HEVC video, the formats both headsets read directly.
The same pipeline ships two ways, and the difference matters. Cloud Conversion runs in the browser: you upload an .mp4 or .mov, the service splits the video into scenes, estimates depth and generates stereo on Depthify.ai's servers, then emails you when the production-quality result is ready. Depthify.ai Desktop is a macOS application for Apple Silicon that does everything locally and works offline. It accepts .jpg, .png, .mp4 and .mov, and produces not only spatial files but depth maps, stereo pairs and 3D meshes. It lets you pick among four named monocular depth models, Depth Anything Small, Depth Anything Base, dpt-hybrid-midas and dpt-large; inspect the prediction in a three.js viewer that projects the image onto a mesh; check the stereo result as an anaglyph or a left and right pair; tune the strength of the depth and stereo effects; and supply your own depth map instead of the estimated one.
The limits are stated plainly. Conversion runs one way only, since 3D back to 2D is not supported. Quality varies with the source material, something the vendor acknowledges rather than hides. There is no API and no mobile application, and the desktop build requires Apple Silicon.
Cloud work is billed purely by volume, from USD 1.50 per minute of 1080p video, with no subscription. The service is young: the domain dates from February 2024, and the product left beta on 28 April 2024.
What it does
- Convert 2D videos in .mp4 or .mov into MV-HEVC spatial video for Apple Vision Pro and Meta Quest
- Convert 2D photos in .jpg or .png into spatial HEIC images
- Generate a depth map from any RGB image or video with a monocular depth network
- Build a left-eye and right-eye stereo pair from an RGB image and its depth map
- Export 3D meshes and inspect them in a three.js viewer
- Process batches through bulk upload, available on every pricing tier
- Run the whole pipeline offline on an Apple Silicon Mac, with nothing uploaded
When to use Depthify.ai / When not to
A quick filter to help you decide if Depthify.ai is the right fit.
When to use Depthify.ai
- Video editors and post-production teams porting an existing 2D catalogue to Apple Vision Pro or Meta Quest
- Creators handling sensitive footage, since the desktop app converts entirely offline on their own machine
- VFX artists and 3D animators who want depth maps, stereo pairs and 3D meshes, not just a finished spatial file
- Studios with volume to process: bulk upload is on every tier, and Enterprise adds 8K, upscaling, segmentation and fisheye undistortion
- Freelancers and small teams selling their output, since a commercial licence is included from the first paid tier
When not to use Depthify.ai
- Developers looking for an API, as none is published and there is no developer documentation
- Windows, Linux and Intel Mac users, because the desktop app requires an Apple Silicon Mac
- Anyone hoping to go the other way, since converting a 3D video back to 2D is explicitly unsupported
- Buyers with a procurement checklist: no DPA, no subprocessor list, no stated GDPR compliance, no published legal entity
- People who want to try before paying, as no free plan and no free trial are announced anywhere on the site
How to use Depthify.ai
A typical end-to-end flow, from setup to results.
- Choose your route: cloud conversion handles video on Depthify.ai's servers, while the desktop app keeps everything on your own Mac
- For the cloud route, sign in at the login step of the conversion flow
- Upload an .mp4 or .mov file and confirm the job
- Wait for the email telling you the converted file is ready
- For the desktop route, download the macOS build for Apple Silicon from the homepage or the footer
- Open a .jpg, .png, .mp4 or .mov file in the application
- Pick one of the four monocular depth models, or supply your own depth map alongside the source file
- Inspect the predicted depth in the three.js viewer, which projects the image onto a 3D mesh
- Adjust the stereo and depth strength, and check the result as an anaglyph or a left and right stereo pair
- Export to HEIC or MV-HEVC and view it on an Apple Vision Pro or Meta Quest; for 8K or a custom pipeline, ask about the Enterprise tier through the contact form
Pros & Cons
Pros
- The desktop application runs entirely offline and, according to the privacy policy, stores only metadata and never the image itself
- The depth models are named and linked publicly, so the technical chain can be inspected rather than taken on trust
- Pricing is published per minute and per frame, with no subscription and no commitment
- A commercial licence is included from the cheapest paid tier
- Content produced by the desktop app is unencumbered: any purpose, including commercial, with no attribution
- Outputs go well beyond the finished spatial file, with depth maps, stereo pairs and 3D meshes available
- You can bring your own depth map, and bulk upload is offered on every tier
Cons
- No public API and no developer documentation
- The desktop application is macOS Apple Silicon only, with no Windows, Linux or mobile version
- No free plan and no free trial are announced, and the desktop app's price is never stated
- No legal entity, postal address or named founder is published anywhere on the site
- No DPA, no subprocessor list, no Article 27 EU representative and no claimed GDPR compliance
- Retention is described only as a reasonable duration, with no figure, and uploaded content may be deleted without notice
- Liability is capped at the lesser of six months of fees or USD 50, and policy changes are declared retroactive
Pricing & Plans
There is no free plan and no free trial. Depthify.ai charges for cloud conversion strictly by volume, and the entry point is the 1080p tier at USD 1.50 per minute of video at 30 frames per second, equivalent to USD 0.00083333 per frame. The 4K tier is charged at USD 3.00 per minute, or USD 0.00166667 per frame. Enterprise pricing is quoted on request through the contact form. The macOS desktop application is offered as a direct download and no price is stated for it on the site.
- 1080p video
- bulk upload
- commercial licence
- 4K video
- bulk upload
- commercial licence
- 1080p
- 4K or 8K video
- bulk upload
- upscaling
- segmentation
- fisheye undistortion
- commercial licence
Data, GDPR & hosting
A consolidated view of how Depthify.ai handles your data.
GDPR overview
Depthify.ai never claims GDPR compliance for itself. The only "GDPR & CCPA compliant" wording on the site describes Umami, the analytics package it uses, and must not be read as a statement about the service. The concrete elements are thin. The privacy policy, effective 24 April 2024, grants access, rectification and deletion on request through team@depthify.ai, with a promise to make reasonable efforts to respond in accordance with applicable laws. Nothing else is stated: no legal basis for processing, no retention period, no international transfers or standard contractual clauses, no Article 27 EU representative, no data protection officer, no DPA and no subprocessor list, even though Cloudflare, Vercel, Sentry and Umami are named as providers. The surrounding framework is visibly American, with DMCA notices under 17 U.S.C. § 512(c) and a California Civil Code § 1542 waiver.
Who owns the data?
You keep your content, and the terms make you solely responsible for it: you must be the creator of the source material or hold sufficient intellectual property rights over it. What Depthify.ai does with it depends on the route. The desktop app never stores the image or video itself, only metadata such as size, format and processing time. Cloud conversion may store the video and all of its metadata on the vendor's servers. Output from the desktop app is yours outright, usable for any purpose without attribution; cloud output requires the commercial licence bundled with the paid tiers. Depthify.ai keeps no backups and may remove uploaded content at any time without notice.
Reuse rights
Reuse depends on which product made the file. For the desktop application, Depthify.ai states that it makes no claim to the content produced and that you are free to use it for any purpose, including commercial purposes, without any attribution: no permission is needed. For cloud conversion the position is narrower, since the terms forbid commercial use of the generated videos without a commercial licence, which is bundled into the 1080p, 4K and Enterprise tiers. In both cases you warrant that you created the material or hold adequate rights to process it, and the prohibited-content rules still apply. One asymmetry is worth noting: any feedback or suggestion you send the vendor is assigned to them outright, and is treated as non-confidential.
Data retention & training
Hosting summary
Depthify.ai never states where data is hosted. The privacy policy, the terms and the cookies policy name no country and no region. Cloud conversions are described only as running on Depthify.ai's secure servers, and the privacy policy names Cloudflare and Vercel among the providers it shares usage information with. Nothing identifies the jurisdiction that would govern a stored video. The desktop application is the clear case: all processing happens on your own machine, it works offline, and nothing is uploaded. Technical signals point to the United States without proving anything about storage. The site resolves to 76.76.21.22 on Vercel infrastructure, geolocated in the United States but flagged as anycast, so the address identifies a content delivery edge rather than a data centre; the domain is registered through NameCheap, Inc., a United States registrar. Anyone with a data residency requirement should ask the vendor directly, or use the desktop application, where the question does not arise.
Things to keep in mind
Risks and trade-offs to weigh before adopting Depthify.ai.
- The only published identity is the brand name: no legal entity, no postal address and no named founder, which leaves little recourse if something goes wrong
- A single address, team@depthify.ai, handles legal, privacy and support alike, so expect no dedicated escalation path
- The cookies policy announces advertising cookies and Google Analytics while the privacy policy names only Vercel Analytics, Umami and Sentry: the two documents contradict each other
- The terms grant a licence exclusively for personal and non-commercial use even though the paid tiers sell a commercial licence, so read that tension before relying on the output commercially
- Cloud uploads and all their metadata may be stored with no stated retention period, the vendor keeps no backups and may delete content without notice: keep your own masters
- Sentry records session replays, so your interactions with the service can be captured well beyond simple page counting
- Depth estimation is an inference, not a measurement: the vendor admits quality varies by content, and treating a generated depth map as accurate 3D data can quietly mislead downstream work
Setup & Integrations
Technical difficulty
Low for the cloud route: sign in, upload an .mp4 or .mov, confirm, and wait for the email. There is nothing to install and no key to configure. The desktop application is a direct download for Apple Silicon Macs, with no server setup and no API. Its controls ask a little more, since choosing among four depth models and tuning the stereo and depth effects assumes you know what a depth map and stereo disparity are. Hardware is the real requirement: an Apple Silicon Mac for the app, and a Vision Pro or Quest to view the result.
Deployment
Supported languages
Behind Depthify.ai
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What does Depthify.ai actually do?
What is the difference between the cloud service and the desktop app?
How much does it cost?
Which file formats can I convert?
Which depth models does the desktop application use?
Can I use the results commercially?
Can Depthify.ai convert a 3D video back into 2D?
What happens to the files I upload?
Is there an API or a mobile application?
How do I contact Depthify.ai?
Should you pick Depthify.ai?
Depthify.ai does one thing, and says clearly how it does it. The three-stage pipeline of monocular depth, stereo disparity and MV-HEVC or HEIC encoding is spelled out on the homepage, and the four depth models behind the desktop application are named and linked rather than hidden behind marketing language. In a category where most tools describe results and nothing else, that transparency counts.
Two things stand out. The desktop application processes everything locally and works offline, and the privacy policy confirms it keeps only metadata, never the image itself, which is a genuine answer for anyone handling material they cannot upload. And the pricing is plain arithmetic: USD 1.50 per minute of 1080p video, USD 3.00 for 4K, published down to the frame, with no subscription and a commercial licence included from the first paid tier.
The weaknesses are institutional rather than technical. No legal entity, postal address or named founder appears anywhere. A single mailbox covers legal, privacy and support. There is no DPA, no subprocessor list and no claimed GDPR compliance; retention is only a reasonable duration; liability is capped at fifty dollars; and policy changes are declared retroactive. The cookies policy and the privacy policy even disagree on which analytics are in use. For an individual creator these are footnotes. For a company with a procurement process they are blockers.
The practical verdict follows that split. If you can work in the desktop application on an Apple Silicon Mac, you get local processing, useful intermediate outputs and unencumbered rights over what you produce. If you need the cloud, you are trusting a young service, whose domain dates from February 2024, with your footage under thin contractual terms. Keep your own masters, and price the risk accordingly.
- Choosing a selection results in a full page refresh.
- Opens in a new window.