Krea

Creates real-time AI image generation with instant visual feedback as users type or draw. Differentiates with real-time generation; mentioned in design tool comparisons for stunning visual output.

Real-time generation enables new creative workflows where users see results instantly while prompting or sketching.

krea.ai Foundation Models - Imageprivate

At a glance

Total raised

$83M

last round series b

Headcount

51-200

approximate · public range

Headquarters

—

Founded

—

Status

Private

Foundation Models - Image

Latest valuation

$500M

Series B · 2025

For subscribers

Subscribe to see the full profile

Including Funding rounds, Bull / Bear thesis, Stock + earnings, Roster changes, Patents, News, and Open roles.

Founding

50% off

/month

94 of 100 spots left

Full

$10

/month

Available once all 100 Founding Member spots are claimed.

Get full access

Already subscribed? Sign in →

Funding history

Total raised

$83M

Rounds tracked

Latest stage

Series B

Latest valuation

$500M

Date	Stage	Amount	Valuation	Lead investors
Apr 7, 2025	Series B	$47M	$500M	Bain Capital Ventures
Jun 1, 2023	Series A	$33M	—	Andreessen Horowitz
Jan 1, 2023	Seed	$3M	—	Pebblebed
Jun 1, 2022	Pre-Seed	—	—	Abstract Ventures

Founder backgrounds

Diego Rodriguez

CTO

Met co-founder Victor Perez as an undergraduate student in Barcelona studying audiovisual systems engineering in the early 2010s. Later worked for Joel Simmons on ArtBreeder, one of the earliest generative AI applications. Won a fellowship from the King of Spain to attend Cornell for graduate studies in computer science, design, psychology, and machine learning, but dropped out when Perez approached him with an early idea for Krea. Founded Krea in 2022 via the HF0 residency program. Rodriguez combines deep expertise in generative AI and creative tools with a relentless focus on product craft—publicly emphasizing that Krea's tools are not made by the Krea team alone, but shaped by human years spent honing design sensibilities and rejecting easier compromises.

Victor Pérez

CEO

Cornell University alumnus who spent early career as Head of Machine Learning Research at Plyzer Intelligence in Barcelona developing graph neural networks and knowledge graphs for e-commerce, then worked as Machine Learning Engineer in Berkeley. Built prosepainter.com and published "Sculpting with Words" at NeurIPS before co-founding Krea AI in March 2022. Met co-founder Diego Rodriguez as students in Barcelona a decade prior; famously spent one day at Cornell before dropping out to pursue early version of Krea. Alumnus of Recurse Center and collaborated on creative projects with NYC Media Lab and CHANEL. Demonstrates persistent conviction in direct human-AI collaboration for creative work—a bet against commoditizing creativity as pure generation.

AI bull / bear

Bull case

Krea's real-time image generation represents a genuine workflow innovation that collapses the feedback loop between creative intent and output. Traditional text-to-image models require users to submit prompts, wait for generation, then iterate—Krea eliminates latency by rendering results as users type or sketch. This "live canvas" experience unlocks new use cases in design prototyping, storyboarding, and iterative concepting where speed matters more than perfect fidelity. The company's positioning in design tool comparisons suggests early product-market fit among creative professionals who value iteration velocity. If Krea can maintain its technical edge in low-latency inference—through model distillation, edge deployment, or proprietary architectures—it could own the real-time niche even as foundation model giants like Midjourney and DALL-E dominate batch workflows. The lack of disclosed funding may indicate bootstrapping or quiet backing, avoiding the valuation pressure that forces premature monetization. Real-time generation could become table stakes for next-gen design tools, positioning Krea as an acquisition target or platform play.

Bear case

Krea faces existential commoditization risk. Real-time generation is an inference optimization problem, not a durable moat—OpenAI, Stability AI, and Midjourney all have the capital and talent to replicate low-latency workflows once they prioritize it. Adobe's Firefly integration into Photoshop and Illustrator already embeds generative AI into incumbent design suites; if Adobe adds real-time rendering, Krea's standalone value collapses. The undisclosed funding history is a red flag. Either the company hasn't raised institutional capital (limiting its ability to scale infrastructure and compete on model quality) or it's operating in stealth with minimal traction. Real-time generation is compute-intensive, requiring expensive GPU infrastructure that's hard to monetize at scale without deep pockets. Without a disclosed customer base, revenue model, or competitive technical advantage beyond speed, Krea risks becoming a feature rather than a platform. Foundation model leaders can afford to subsidize real-time as a loss leader; Krea likely cannot.

Generated by claude-sonnet-4-5-20250929 · May 1, 2026Refreshes quarterly · see disclosures

In the news

Krea · last 30 days

No articles ingested yet for Krea. Once the hourly news pipeline is live, every article the classifier tags as mentioning this company appears here with its one-line AI summary and sentiment.

Watch

Leadership

Victor Pérez

CEO

in 𝕏

Diego Rodriguez

CTO

in 𝕏

Products

Real-Time Image Generation
Market-leading real-time image generation tool that produces photorealistic images in under 50 milliseconds as users type or draw. Offers access to 40+ AI models including proprietary Krea 1, Flux, Imagen, and others, with intuitive brush and shape-based controls alongside traditional text prompts. Designed for designers, artists, and creatives to rapidly iterate visual concepts without the typical submit-wait-regenerate cycle. Supports instant visual feedback enabling seamless creative exploration.
Video Generation
Comprehensive video creation suite providing access to advanced models including Veo 3, Kling, Hailuo, Wan, and Runway. Enables users to generate viral videos from text prompts, animate static images, convert keyframe sequences into video content, and produce professional-quality short-form videos. Includes timeline-based editing interface familiar to filmmakers and video editors, supporting duration adjustment, frame reordering, and multi-shot generation with high coherence and cinematic camera control.
Image Upscaling & Enhancement
Professional-grade image enhancement tool supporting resolution upscaling up to 22K pixels using multiple upscaling models. Restores old videos, turns low-resolution phone captures into professional footage, and adds ultra-fine textures to portraits. Includes specialized enhancement models to improve blurry photos, restore archived content, and apply advanced filtering. Features video upscaling up to 8K with frame interpolation to 120fps, supporting slow-motion and restoration workflows.
Generative Image Editing
AI-powered image editor with 10+ specialized models enabling precise control over image modifications. Users can add or remove objects, merge images, adjust lighting and expressions, change specific regions with separate prompts through annotation tools, and expand scenes using generative AI. Offers mask-based editing with fine-grained AI control and layered editing capabilities. Supports multiple simultaneous edits through annotation interface allowing complex creative adjustments within single pass.
3D Generation & Training
Creates 3D objects and meshes from text descriptions or converts 2D images into 3D models for game development and product visualization. Includes custom model training (LoRA fine-tuning) allowing users to upload images of specific characters, products, or styles to generate consistent custom variations on demand. Supports training on custom datasets for brand-specific aesthetics, character consistency, and artistic style reproduction across all generation tools.

Open roles

View all careers ↗