Ideogram · The open-weight model that finally gets typography right.

Ideogram 4.0

Design-Ready Images, Text That Reads

Ideogram 4.0 is the vendor's first open-weight foundation model, a 9.3B-parameter Diffusion Transformer built for flawless in-scene typography, deterministic layout, and native 2K photoreal design output.

9.3B
Parameters (DiT)
2048px
Native resolution
0.97
English OCR score
3
Speed-quality tiers
What's new

Design-Ready Images, Text That Reads

01

Best-in-class text rendering

Renders headlines, logos, and paragraph copy legibly inside the scene, scoring 0.97 on English OCR accuracy and leading every other open-weight model on text benchmarks.

02

JSON layout control

Go beyond plain prompts with a structured JSON schema: bounding-box coordinates place each element, and hex palette conditioning locks exact brand colors.

03

Native 2K, open weights

Generates up to 2048px per side and ships public fp8 and nf4 checkpoints on Hugging Face and GitHub for local inference.

04

Three speed-quality tiers

Choose Turbo for fast drafts, Balanced for everyday work, or Quality for maximum-fidelity final renders.

Overview

Model at a glance

Ideogram 4.0, released on June 3, 2026, is the Toronto-based lab's most capable text-to-image release and its first open-weight foundation model. At its core is a 9.3 billion-parameter single-stream Diffusion Transformer trained from scratch, where text and image tokens share the same projections rather than running through separate branches. It pairs this with Qwen3-VL-8B-Instruct as its text encoder, concatenating hidden states from 13 intermediate layers so the model reads prompts with the semantic depth of a full vision-language model.

The release targets the parts of image generation that matter most for real design work: typography that stays legible inside the scene, layouts you can control precisely, and clean 2K photoreal output. A structured JSON prompting format lets you define bounding boxes for element placement, assign hex color palettes, and separate the text string from its visual styling for each element in the frame. Ideogram ships public fp8 and nf4 checkpoints for local inference, while the hosted API and partner endpoints cover production-scale and commercial use.

Model family
Ideogram 4.0
Modality
Text-to-image
Parameters
9.3 billion
Architecture
Single-stream Diffusion Transformer
Text encoder
Qwen3-VL-8B-Instruct
Max resolution
Native 2K (2048px/side)
Weights
Open (fp8 & nf4 checkpoints)
Release date
June 3, 2026
Capabilities

Key features

In-scene typography that reads

Ideogram 4.0's headline strength is text rendering. It reliably places legible words, headlines, and even paragraph-length copy inside a generated image, scoring 0.97 on English OCR accuracy and topping open-weight text-rendering benchmarks ahead of much larger models.

Structured JSON prompting

Instead of relying on a single natural-language string, you can pass a JSON schema that defines regions, assigns content to them, and controls styling per element. This gives layout-level determinism that plain-text prompts on most models cannot match.

Bounding-box layout control

Specify bounding-box coordinates to pin logos, headings, and product elements exactly where you want them, turning composition from a guessing game into a spec you define up front.

Hex palette conditioning

Feed exact hex colors into the prompt to constrain the output palette, keeping generated designs on-brand without color drift across a batch.

Native 2K photoreal output

The model generates natively up to 2048px per side across a range of aspect ratios, producing print- and screen-ready design assets without a separate upscaling pass.

Open weights for local inference

As Ideogram's first open-weight model, 4.0 publishes its weights and inference code on Hugging Face and GitHub, with quantized fp8 and nf4 checkpoints that make local and self-hosted generation practical.

Three speed-quality tiers

Turbo trades steps for speed on quick iterations, Balanced covers everyday production, and Quality maximizes fidelity for final renders, letting you dial cost and detail per job.

Specs

Technical specifications

Model architecture

Parameters9.3 billion
ArchitectureSingle-stream Diffusion Transformer (trained from scratch)
Text encoderQwen3-VL-8B-Instruct (13 intermediate layers concatenated)
ModalityText-to-image

Output & tiers

Max resolutionNative 2K (up to 2048px per side)
TiersTurbo, Balanced, Quality
FormatsPNG, WEBP (with transparency support)
Aspect ratiosMultiple supported

Control & prompting

Prompt formatNatural language + structured JSON schema
Layout controlBounding-box coordinates per element
Color controlHex palette conditioning
Per-element controlSeparate text string and visual styling

Distribution

WeightsOpen (Hugging Face + GitHub)
Checkpointsfp8 and nf4 quantized variants
LicenseIdeogram 4 Non-Commercial Model Agreement (weights); commercial via API/subscription
ReleasedJune 3, 2026
In practice

Use cases

Posters and marketing graphics

Generate finished posters, ads, and social graphics with headlines and body copy rendered legibly in-scene, no manual text overlay required.

Logo and typographic concepts

Explore wordmarks, badges, and typographic treatments where accurate letterforms and spelling are non-negotiable.

On-brand asset batches

Use hex palette conditioning and bounding-box layout to produce large sets of visuals that stay consistent with a brand's colors and composition rules.

Packaging and product mockups

Place labels, product names, and callouts precisely with JSON layout control to prototype packaging and merchandise designs.

Local and self-hosted generation

Run the open fp8 or nf4 checkpoints on your own hardware for private, offline, or high-volume design pipelines.

Honest look

Current limitations

Non-commercial open weights

The published weights are released under a non-commercial agreement. Commercial use runs through Ideogram's subscription, developer API, or hosted partner endpoints rather than the open checkpoints.

Text-to-image only

Ideogram 4.0 is a text-to-image foundation model. It does not generate video or audio, and image-editing or inpainting capabilities are separate from base generation.

JSON prompting has a learning curve

The structured JSON schema that unlocks precise layout and color control is more involved than plain-text prompting and takes experimentation to use well.

Vendor-reported benchmarks

Figures such as the 0.97 OCR score, layout mIoU, and DesignArena rankings are largely reported by Ideogram; independent results may vary by prompt, tier, and settings.

FAQ

Frequently asked questions

What is Ideogram 4.0?
Ideogram 4.0 is Ideogram's first open-weight text-to-image foundation model, released June 3, 2026. It is a 9.3-billion-parameter single-stream Diffusion Transformer built for best-in-class in-scene typography, deterministic layout, and native 2K output.
What makes Ideogram 4.0 good at text?
It was designed around rendering readable text inside images. It scores 0.97 on English OCR accuracy and leads open-weight models on text-rendering benchmarks, reliably producing legible headlines, logos, and paragraph copy.
What is JSON prompting in Ideogram 4.0?
Alongside a natural-language prompt, you can pass a JSON schema that defines bounding-box regions, assigns content to them, applies hex color palettes, and separates each text element's string from its styling, giving layout-level control over the output.
Are the Ideogram 4.0 weights really open?
Yes. Ideogram published fp8 and nf4 checkpoints and inference code on Hugging Face and GitHub. The weights are under a non-commercial agreement, so commercial use goes through Ideogram's subscription, developer API, or hosted partners.
What resolution and tiers does Ideogram 4.0 support?
It generates natively up to 2048px per side across multiple aspect ratios, with three inference tiers, Turbo, Balanced, and Quality, to trade speed against fidelity.
How does Ideogram 4.0 compare to larger open models?
Despite its relatively compact 9.3B size, Ideogram reports it delivers the best text rendering of any benchmarked open-weight release, ahead of larger models, and ranked first among open weights on the DesignArena leaderboard.
Can I use Ideogram 4.0 on SharkFoto?
Not yet. Ideogram 4.0 support on SharkFoto is coming soon. In the meantime you can create with SharkFoto's currently available AI image tools and effects, and this model will be added to the lineup as integration completes.

Ideogram 4.0 is coming to SharkFoto

Design-ready images with text that actually reads are on the way. Ideogram 4.0 support is coming soon to SharkFoto. Explore our available AI image tools and effects today, and be ready to generate poster-perfect typography the moment it lands.

Try Ideogram 4.0 now