Recraft V3
Recraft V3, codenamed red_panda before launch, is the model that put Recraft on the map when it topped the Artificial Analysis text-to-image leaderboard in October 2024. It is still sold at $0.04 per image and keeps two things V4.1 dropped on fal: a catalogue of about seventy named styles across realistic, digital illustration and vector, and an image-to-image endpoint with a strength dial.
Anvisha Pai, Co-founder & CEO, ModaTested Recraft · released
›What we asked for
A self-portrait of Recraft V3, an AI image model known for styles, as an isometric 3D icon set: nine small isometric objects on a pastel grid, each a different art style from photoreal to pixel art to line art, soft shadows, a tiny label 'V3' on one tile.
fal-ai/recraft/v3/text-to-image
image_size: {"width":1360,"height":1024}, style: realistic_image, num_images: 1
returned 1365 by 1024 pixels
At a glance
- Price
- $0.04 per raster image, $0.08 for vector styles.
- Styles
- About seventy named styles in three families. The ladder below shows the same prompt as a photo, a digital illustration and a vector.
- Image to image
- A strength dial from 0 to 1 says how far from the source to stray.
- Brand colours
- Pass preferred colours, as with V4.1.
- Age
- Two generations old; V4.1 is the current model and cheaper on fal.
Other Recraft models: Recraft V4.1. Compare all 2.
Anvisha's take
V3 is worth keeping for two things, and the ladder shows both. Ask for realistic_image and it interprets a poster prompt as a photograph of someone holding a poster, which is either a delightful literalism or a trap. Ask for digital_illustration and you get a proper screen-print; ask for vector_illustration and you get clean flat art at double the price. The image-to-image edit regenerated the whole scene around the dripper. Cheap, fast, and a style catalogue no other model has.
Recraft V3 real output: the same 12 prompts every model gets
Every model on this site gets the same 12 prompts, so you can compare like with like. These are the pictures Recraft V3 returned, one try each, nothing re-rolled or retouched. Generated on . The 12 pictures took 5 to 14 seconds each and cost $0.48 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/recraft/v3/text-to-image
image_size: {"width":1024,"height":1360}, style: realistic_image, num_images: 1
returned 1024 by 1365 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/recraft/v3/text-to-image
image_size: {"width":1024,"height":1024}, style: realistic_image, 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/recraft/v3/text-to-image
image_size: {"width":1024,"height":1360}, style: realistic_image, num_images: 1
returned 1024 by 1365 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/recraft/v3/text-to-image
image_size: {"width":1024,"height":1024}, style: realistic_image, 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/recraft/v3/text-to-image
image_size: {"width":1360,"height":1024}, style: realistic_image, num_images: 1
returned 1365 by 1024 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/recraft/v3/text-to-image
image_size: {"width":1024,"height":1360}, style: realistic_image, num_images: 1
returned 1024 by 1365 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/recraft/v3/text-to-image
image_size: {"width":1360,"height":1024}, style: realistic_image, num_images: 1
returned 1365 by 1024 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/recraft/v3/text-to-image
image_size: {"width":2016,"height":864}, style: realistic_image, num_images: 1
returned 2048 by 1024 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/recraft/v3/text-to-image
image_size: {"width":1024,"height":1360}, style: realistic_image, num_images: 1
returned 1024 by 1365 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/recraft/v3/text-to-image
image_size: {"width":1024,"height":1024}, style: realistic_image, 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/recraft/v3/text-to-image
image_size: {"width":1360,"height":1024}, style: realistic_image, num_images: 1
returned 1365 by 1024 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/recraft/v3/image-to-image
style: realistic_image, strength: 0.5
returned 1024 by 1024 pixels
How the cost under each picture was worked out: fal's flat per-image price for Recraft V3: $0.04, or $0.08 for vector styles (fal model page, read 2026-09-09).
Recraft V3 settings that matter
Style family
Three styles, three different pictures from the same words. realistic_image turned the poster prompt into a photograph of a person holding the poster, text correct and date awkwardly broken; digital_illustration produced a genuine screen-print, the best poster of the three; vector_illustration produced flat line art with a lowercase 'june' and cost double. On the pizza, vector is a charming menu illustration and realistic is a photo. Pick the style for the medium you are designing for, and expect realistic to take 'poster' literally.
Text in an image
›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/recraft/v3/text-to-image
image_size: {"width":1024,"height":1360}, style: realistic_image, num_images: 1
returned 1024 by 1365 pixels
Food photography
›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/recraft/v3/text-to-image
image_size: {"width":1024,"height":1024}, style: realistic_image, num_images: 1
returned 1024 by 1024 pixels
Recraft V3 pricing
| Where | Unit | Price | Source |
|---|---|---|---|
| Recraft API | Raster image 40 API units; vector is 80 units. | $0.04 | Recraft API pricing, read |
| fal | Raster image Vector styles are $0.08. | $0.04 | fal model page, read |
Is Recraft V3 free?
Recraft's own app has a free plan with a daily credit allowance, but images made on it are owned by Recraft, shown publicly in the community gallery and not licensed for commercial use. Paid plans grant full ownership. The API has no free tier.
- •Recraft app, free: daily credits (coverage puts it at about 50 images a day), outputs public and non-commercial.
- •Recraft app, paid: from about $12 a month, with ownership and commercial rights.
- •API: billed in API units, 1,000 units to the dollar; V4.1 is 35 units per raster image.
- •fal and other hosts: pay per image.
Recraft pricing, read
Where to use Recraft V3
- Recraft app · Recraft Studio, with vector output, brand styles and a free plan.
- Recraft API · Billed in API units; raster and vector variants.
- fal · Flat per-image pricing.
- Moda · Available inside the Moda design agent as Recraft V3.
Generate with Recraft V3 in Moda
Pick Recraft V3 in the Moda design agent, then edit the result as vectors, text and layers rather than a flat image.
Open Moda →What we noticed running Recraft V3
- •Text is correct in every style, but the realistic style reads 'poster' as a photograph of a person holding one, with the date split across two lines. See the ladder. See the text in an image picture.
- •Image-to-image is not an edit. At strength 0.5 the dripper came back green with a brass handle and a whole new kitchen behind it, bowl, jug and plant included. See the edit an existing image picture.
- •Vector styles cost double: $0.08 against $0.04, and on Recraft's own API they return real SVG.
- •Fast: 5 to 14 seconds per picture, median 7.
- •Sizes are Recraft's presets: 1024 by 1365 at 3:4, 2048 by 1024 at 21:9.
- •The self-portrait added a 'V3' tile to the icon set without being asked, which is the kind of on-brief invention that is fine here and a liability on a client job.
Recraft V3 limits and API parameters
| Limit | Value |
|---|---|
| Output size | Presets or any width and height. fal API reference, read |
| Edits | Image-to-image with a strength dial, not a true overlay edit; see the edit sample. fal API reference, read |
›Every setting, for developers
fal (fal-ai/recraft/v3/text-to-image)
fal API reference, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| prompt | string | Any text Required. | |
| image_size | enum or object | square_hd, square, portrait_4_3, portrait_16_9, landscape_4_3, landscape_16_9, or {width, height} | square_hd |
| style | enum | realistic_image, digital_illustration, vector_illustration and about seventy sub-styles | realistic_image |
| colors | list of RGB | Preferred colours | |
| style_id | string | A custom style id | |
| image_url | URL | Source image Image-to-image endpoint only. | |
| strength | number | 0 to 1 Image-to-image endpoint. | 0.5 |
What Recraft V3 will not draw
Recraft's terms bar content that exploits children, non-consensual intimate imagery or deepfakes of real individuals, content that falsely depicts a real person in a political context, material that promotes terrorism or violence against protected groups, and removing AI-disclosure watermarks or metadata. Two rules matter for design work: images made on the free plan are owned by Recraft, public in the community gallery and not licensed for commercial use, and generated assets may not be used to train AI models.
Recraft terms of service, read
Alternatives to Recraft V3
Frequently asked questions
What is Recraft V3?
Recraft V3 is the October 2024 model that topped the Artificial Analysis leaderboard, still sold at $0.04 per image with about seventy named styles and an image-to-image endpoint.
Which Recraft style should I use?
The ladder above shows the same two prompts as realistic_image, digital_illustration and vector_illustration. Vector costs double and returns SVG on Recraft's own API.
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.
