HiDream-I1 Full
HiDream-I1 is HiDream.ai's open-weight image model, released on 7 April 2025 under the MIT licence with 17 billion parameters and three variants: Full at 50 steps, Dev at 28 and Fast at 16. It posted strong scores on the prompt-following benchmarks of the time and was one of the largest truly open image models. On fal the Full variant costs $0.05 a megapixel and takes over a minute an image, which is the trade: permissive weights and a big model against speed and a roster that has moved on.
Anvisha Pai, Co-founder & CEO, Voyager
HiDream.ai · released
›What we asked for
A self-portrait of HiDream-I1, an open-weight image model with seventeen billion parameters, as a lucid dream: a bedroom whose ceiling opens onto a galaxy, a paper boat sailing on the bedsheets, soft cinematic light.
fal-ai/hidream-i1-full
image_size: {"width":1360,"height":1024}, output_format: png, num_images: 1
returned 1168 by 880 pixels
At a glance
- Price
- $0.05 a megapixel on fal, rounded up; the eleven prompts that ran cost $0.95.
- Open weights
- MIT licence, 17 billion parameters, Full, Dev and Fast variants on Hugging Face.
- Speed
- A 104-second first call while the model loaded, then about 13 seconds an image at fifty steps.
- Sizes
- fal's presets or custom pixels; our requested sizes came back slightly different.
- Editing
- None on this endpoint; HiDream-E1 is the separate editing model.
Anvisha's take
HiDream-I1 Full shows its age next to the 2026 roster and still has a few things to say. The pizza and the placement test are as good as anything here, the woodblock lighthouse is a handsome mid-century print, and the pour-over portrait is natural. The failures are in text and data: the poster's title lines overlap, the logo grew a 'REPAIR SHOP' line nobody asked for, and the chart became a two-column table with invented numbers. The first image took 104 seconds while the model loaded and the rest about 13 each, and the returned sizes did not match the request. The reason to use it is the MIT licence, not the output.
HiDream-I1 Full real output: the same 12 prompts every model gets
Every image model on this site gets the same 12 prompts, so you can compare like with like. These are the pictures HiDream-I1 Full returned, one try each, nothing re-rolled or retouched. Generated on . The 11 pictures took 13 to 104 seconds each and cost $0.95 in total.
Spelling, kerning and hierarchy of three lines of set text.
›What we asked for
A concert poster that reads 'MODA SUMMER SESSIONS' in large bold type, with the subtitle 'Friday 12 June, 8pm, Roundhouse' and a smaller line 'Tickets at the door'. Risograph style, two colours, tangerine and navy, grainy ink texture.
fal-ai/hidream-i1-full
image_size: {"width":1024,"height":1360}, output_format: png, num_images: 1
returned 880 by 1168 pixels
Catalogue lighting, a clean seamless background and a believable contact shadow.
›What we asked for
Studio product photograph of a matte black ceramic pour-over coffee dripper on a seamless white background, soft top light, subtle contact shadow, centred, catalogue style.
fal-ai/hidream-i1-full
image_size: {"width":1024,"height":1024}, output_format: png, num_images: 1
returned 1024 by 1024 pixels
A natural face, correct finger count and hands holding an object.
›What we asked for
Close-up portrait of a woman in her 40s with short grey hair, laughing, holding a red ceramic mug with both hands, natural window light, 85mm lens, shallow depth of field.
fal-ai/hidream-i1-full
image_size: {"width":1024,"height":1360}, output_format: png, num_images: 1
returned 880 by 1168 pixels
A flat single-colour mark with a legible wordmark and no gradients or photo texture.
›What we asked for
A flat vector logo mark for a bicycle repair shop called 'Spoke & Chain': a single-colour icon combining a chain link and a wheel spoke, with the wordmark below in a geometric sans-serif, black on white.
fal-ai/hidream-i1-full
image_size: {"width":1024,"height":1024}, output_format: png, num_images: 1
returned 1024 by 1024 pixels
Many distinct objects and people in one frame without merging or duplicating them.
›What we asked for
A crowded Saturday farmers market seen from slightly above: at least twelve stalls with striped awnings, shoppers carrying tote bags, a dog on a leash, crates of oranges, a musician with a guitar, bunting overhead, late morning light.
fal-ai/hidream-i1-full
image_size: {"width":1360,"height":1024}, output_format: png, num_images: 1
returned 1168 by 880 pixels
A named historical style applied faithfully rather than a generic painterly look.
›What we asked for
A lighthouse on a rocky cliff during a storm, in the style of a Japanese ukiyo-e woodblock print, with visible wood grain texture, flat colour areas and a bokashi gradient sky.
fal-ai/hidream-i1-full
image_size: {"width":1024,"height":1360}, output_format: png, num_images: 1
returned 880 by 1168 pixels
Six labelled bars at the stated heights, in order, with readable axes.
›What we asked for
A clean bar chart titled 'Monthly active users' showing six months, January through June, with the values 120, 150, 180, 170, 220 and 260, labelled axes, white background, one accent colour, no 3D effects.
fal-ai/hidream-i1-full
image_size: {"width":1360,"height":1024}, output_format: png, num_images: 1
returned 1168 by 880 pixels
A 21:9 composition that leaves the left third clear for headline text.
›What we asked for
Website hero banner: a wide panorama of a sunlit modern co-working space with plants, the left third of the frame kept as clear wall space for text, warm neutral palette.
fal-ai/hidream-i1-full
image_size: {"width":2016,"height":864}, output_format: png, num_images: 1
returned 1360 by 768 pixels
Four numbered steps in the given order with their exact wording, plus an icon each.
›What we asked for
An infographic card titled 'How to brew pour-over' with four numbered steps in this order: 1. Rinse the filter, 2. Bloom for 30 seconds, 3. Pour in slow circles, 4. Wait 3 minutes. A simple icon beside each step, muted palette, clean sans-serif type.
fal-ai/hidream-i1-full
image_size: {"width":1024,"height":1360}, output_format: png, num_images: 1
returned 880 by 1168 pixels
Appetising, menu-grade food with correct textures and no melted anatomy.
›What we asked for
Overhead photograph of a margherita pizza on a wooden board, fresh basil leaves, one slice pulled out with the cheese stretching, restaurant menu photography, natural light.
fal-ai/hidream-i1-full
image_size: {"width":1024,"height":1024}, output_format: png, num_images: 1
returned 1024 by 1024 pixels
Exact counts and left-right placement of five named objects.
›What we asked for
A glass of water on a wooden table with exactly three ice cubes in it, a yellow pencil resting across the rim of the glass, a blue book to the left of the glass and a red apple to the right, photographed from table height.
fal-ai/hidream-i1-full
image_size: {"width":1360,"height":1024}, output_format: png, num_images: 1
returned 1168 by 880 pixels
Change one object's colour and add a part while leaving background, light and framing untouched.
›What we asked for
Change the coffee dripper to a deep forest green and add a small brass handle on its right side. Keep the background, lighting and composition exactly as they are. (applied to the product on white result above)
fal-ai/hidream-i1-full
How the cost under each picture was worked out: fal's HiDream-I1 Full rate of $0.05 per megapixel (fal model page, read 2026-09-10), rounded up to the nearest 1024x1024 megapixel as fal states: 2 for the 1024x1360 output; fal does not return the charge.
HiDream-I1 Full pricing
| Where | Unit | Price | Source |
|---|---|---|---|
| fal | Megapixel Rounded up to the nearest megapixel. | $0.05 | fal model page, read |
| Self-hosted | Image Free under MIT; needs a large GPU. | $0.00 | HiDream-ai/HiDream-I1 on GitHub, read |
Is HiDream-I1 Full free?
The weights are free under MIT, and HiDream's Vivago app offers free credits. Hosted endpoints charge per megapixel.
- •Open weights: Full, Dev and Fast on Hugging Face under MIT.
- •Vivago app: free credits from HiDream.ai.
- •fal: per megapixel, no free tier.
Where to use HiDream-I1 Full
- Open weights · MIT weights and inference code.
- Vivago · HiDream.ai's own app.
- fal · $0.05 a megapixel; Dev and Fast variants are separate endpoints.
- Voyager · Runs inside Voyager, the agent for creative work. Private preview by waitlist.
Voyager
Run HiDream-I1 Full inside Voyager
Voyager is an agent for creative work that runs every model in this reference, HiDream-I1 Full included. Brief it, and it picks the model, makes the thing and hands back something you can edit. Create an account to get started.
What we noticed running HiDream-I1 Full
- •Returned a table with two columns headed January and June and numbers that are not the ones in the prompt; no bars at all. See the chart picture.
- •'MODA' and 'SUMMER' overlap on the poster, with the date line correct underneath. See the text in an image picture.
- •Added a 'REPAIR SHOP' line under the wordmark, as both Nano Banana 2 tiers did. See the logo mark picture.
- •The first prompt took 104 seconds and the other eleven about 13 each, so budget for a cold start. Requested sizes came back a little different: 880 by 1168 for a 1024 by 1360 request.
HiDream-I1 Full limits and API parameters
| Limit | Value |
|---|---|
| Speed | About 13 seconds an image on fal at fifty steps once warm; the first call took 104 seconds. fal model page, read |
| Editing | None on this endpoint. fal API reference, read |
›Every setting, for developers
fal (fal-ai/hidream-i1-full)
fal API reference, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| prompt | string | Any text Required. | |
| image_size | enum or {width, height} | square_hd, square, portrait_4_3, portrait_16_9, landscape_4_3, landscape_16_9, or custom pixels | 512x512 |
| num_inference_steps | integer | Steps | 50 |
| guidance_scale | number | Prompt adherence | 5 |
| negative_prompt | string | Any text | |
| num_images | integer | 1 and up | 1 |
| output_format | enum | jpeg, png | jpeg |
| seed | integer | Any integer | |
| enable_safety_checker | boolean | true, false | true |
What HiDream-I1 Full will not draw
The hosted endpoint runs behind fal's safety checker, on by default. The MIT weights carry no usage restrictions beyond the licence, so self-hosted the model has no filter you do not add.
fal API reference, read
Alternatives to HiDream-I1 Full
Frequently asked questions
What is HiDream-I1?
HiDream-I1 is HiDream.ai's open-weight image model from 7 April 2025: 17 billion parameters under MIT, in Full, Dev and Fast variants. Full is on fal at $0.05 a megapixel.
Is HiDream-I1 open source?
Yes, the weights are MIT licensed and on Hugging Face, one of the most permissive licences of any large image model.
Is HiDream-I1 still worth using?
For self-hosting under MIT, yes. Hosted, it is slow and the twelve prompts below show where newer models have passed it.
How this page is made
Prices, limits and settings are read from the linked vendor and host pages on the dates shown. Every picture was generated by us through fal with the request shown under it, and the original files are kept unedited. A generation that failed is shown as a failure. Model and vendor names are used to identify the products; no vendor imagery appears on this page.
