Learn About Stability AI's AI Products
Create a free account to access in-depth lessons on each tool and model.
Start Learning Free📋About Stability AI
Updated June 15, 2026Stability AI is a generative AI company founded in 2019, best known for Stable Diffusion — the most influential open-weight image generation model. Under new CEO Prem Akkaraju (appointed June 2024, former Weta Digital CEO), the company has stabilized after financial difficulties and leadership turmoil following founder Emad Mostaque's departure.
Stable Diffusion 3.5 (October 2024) is the latest image-generation generation, using the MMDiT architecture with improved text rendering. Three variants are available: Large (8.1 billion parameters), Large Turbo, and Medium (2.5 billion, runs on consumer RTX 3080+ GPUs). The Community License allows free use for individuals and businesses under one million dollars in revenue.
In audio, Stability AI's current flagship is Stable Audio 3.0 — a four-model family for music generation. The medium and large variants render compositions up to six minutes and twenty seconds; the small variants top out at two minutes. Three of the four variants ship with open weights (small SFX, small, and medium), while the large model is API or self-hosted under a paid license required above one million dollars in revenue. Stable Audio 3.0's defining structural choice is that its training data is fully licensed via direct partnerships with Warner Music Group and Universal Music Group — a deliberate contrast with rival music-generation tools Suno and Udio, both of which are fighting major-label copyright suits. The licensed-data posture extends a broader Stability AI strategy of competing on legal posture alongside raw model capability.
In November 2025, the UK High Court rejected Getty Images' central copyright claim against Stability AI — the first major UK ruling on copyright and generative AI training. Stability AI generated approximately $50 million in 2024 revenue and received investment from WPP. The multi-modal product line spans image (Stable Diffusion 3.5), audio (Stable Audio 3.0), 3D (SPAR3D), and video (SV4D).
🛠️Products & Tools (3)
The leading open-source image generation model. Run locally, fine-tune on custom data, or use via API. Huge ecosystem of models, LoRAs, and tools.
Stability AI's latest open-source image generation model. Multiple variants for different hardware. Company stabilized under new CEO Prem Akkaraju; Getty Images lawsuit ongoing.
Stability AI's four-model family for music generation. Medium and large variants render compositions up to six minutes and twenty seconds; small variants top out at two minutes. Three of the four variants ship with open weights (small SFX, small, and medium); the large model is API or self-hosted under a paid license required above one million dollars in revenue. Training data is fully licensed via direct partnerships with Warner Music Group and Universal Music Group.