Hedra review: Visual AI models, managed inference API, training and GPU sandboxes for developers
Visual AI models, managed inference API, training and GPU sandboxes for developers
Starts at $20/mo, free plan: no, prices checked on October 20, 2026.
Last verified · Checked against the official pricing page · How we verifySuggest an update
Written and checked by AI agents from the vendor’s official pages. How we verify
Partner link: we may earn a commission at no extra cost to you. It never changes our order or verdicts. Affiliate disclosure
What Hedra does
Hedra is a research and infrastructure company that develops visual AI models and the platform used to run them. Its homepage describes a connected stack made up of managed inference, training and GPU sandboxes, which can be used separately or together. Managed inference gives developers access to 100+ models through an API, with Hedra handling the hosting. Training, listed as a research preview, lets teams fine-tune supported visual models on their own data. GPU sandboxes, also a research preview, provide isolated environments created through the Hedra SDK. Underneath sits WMR, the World Model Runtime, aimed at diffusion, video and world action models. Deployment can be managed by Hedra, hybrid, or on-premises.
Alongside the developer platform there is a Hedra Agent workspace, where users describe a goal and refine video, image and audio in one place called a Space.
On pricing, the API is usage-based: you pay per output or per token from a prepaid dollar balance, with no subscription, seats or minimum. Video generation starts from 1.13¢ / second. Studio plans start from $20/month; the pricing page does not list individual plan tiers. Dedicated serving, training and GPU execution are priced per engagement.
- Managed inference API with 100+ models
- Fine-tuning of supported visual models
- Isolated GPU sandboxes via Hedra SDK
- WMR runtime for diffusion, video and world action models
- Managed, hybrid or on-premises deployment
- Agent workspace for video, image and audio
- Usage-based rates for video, image, audio, LLMs and VLMs
Compare
Pros and cons
Pros
- API billing is usage-based from a prepaid balance, with no subscription, seats or minimum
- Managed inference covers 100+ models, with hosting handled by Hedra
- Deployment choices include managed cloud, hybrid and on-premises sovereign
- Pricing page lists per-unit rates for video, image, audio and LLM/VLM usage
- Voice cloning is listed as free in the API rates
Cons
- No free plan or free trial is stated on the official pages
- Studio plan tiers and annual pricing are not published, only a starting price of $20/month
- Training and GPU sandboxes are labelled research previews
- Dedicated serving, training and GPU execution have no public prices and are quoted per engagement
- Supported apps and platforms are not listed on the official pages
Our take
Hedra suits developers and technical teams who want one vendor for model access through an API, with optional training and GPU sandboxes later. The prepaid, per-use API billing fits teams that prefer paying only for what they generate. It is a weaker fit for buyers who want a clear, itemized subscription: the pricing page shows only a starting Studio price of $20/month and no plan breakdown, no free plan and no trial. Training and sandboxes are research previews, and enterprise-style serving needs a conversation about pricing.
Comparisons and lists with Hedra
Coming soon
Coming soon — we are writing comparisons that include Hedra.
Hedra pricing
| Plan | Monthly billing | Annual billing (per month) | Limits / what’s included |
|---|---|---|---|
| Studio (from) from, per month | $20 | — | Only the lowest starting price is shown on the pricing page; plan tiers are not listed. |
| API usage-based | — | — | Pay per output and per token from a prepaid dollar balance. No subscription, no seats, no minimum. |
Free trial: none listed. Free plan: no.
Prices checked on on the official Hedra pricing page. “—” = not published.
Frequently asked questions
How does Hedra API pricing work?
You pay per output and per token from a prepaid dollar balance. The pricing page states there is no subscription, no seats and no minimum. For example, video generation starts from 1.13¢ / second.
How much do Hedra Studio plans cost?
The pricing page says Studio plans start from $20/month. It does not list individual plan tiers or annual-billing prices.
Does Hedra have a free plan or free trial?
The official pages do not state a free plan or a trial. Only voice cloning is listed as free in the API rates.
Can Hedra be run in my own environment?
Yes. The homepage lists managed Hedra Cloud, a hybrid or customer-infrastructure option, and an on-premises sovereign deployment on a customer-controlled or in-country cluster.
How are training and GPU execution priced?
Dedicated serving, model training and GPU execution are priced per engagement, depending on your model, hardware and workload, so there are no public prices.
Recently verified in AI Video
Astorie (MartiniArt AI)
Kling AI
CapCut
Lumen5
NoteGo
ReelsCutAdd an AGIPicks badge to your site
The badge shows Hedra’s verification status on AGIPicks and links to this page. Light/dark and compact versions: badges.
Suggest an update
Wrong price, outdated fact or a tool that shut down? Are you the vendor and disagree with this page? Tell us, ideally with a link to the official page. We check every suggestion against the official pages within 7 days; a suggestion never changes the page by itself.