Stability AI

Available

Stability AI has 7 models tracked on Modelglass across 3 verticals (Image, Audio, 3D), most recently added 2025-01.

Stability AI has 2 image models (2 current-gen), 3 audio models (1 current-gen, 1 previous-gen, 1 deprecated), and 2 3d models on Modelglass, most recently added 2025-01.

stability.ai ↗

Coverage across verticals

Image 2 models

2 current-gen

Image pricing page ↗

Pioneer of open-weight image generation (Stable Diffusion family, SD3, SDXL). Offers a direct inference API at platform.stability.ai alongside open-weight downloads. Models are also widely available via third-party hosts (Replicate, fal.ai) — those offerings are tracked under those provider slugs. This entry covers direct Stability AI API access only.

Audio 3 models

1 current-gen · 1 previous-gen · 1 deprecated

Audio pricing page ↗

Stability AI offers Stable Audio for music generation via API. Credit-based pricing: 1 credit = $0.01; 25 free credits on signup. Current models: Stable Audio 2.5 (20 credits/$0.20, up to 3 min) and Stable Audio 3.0 (26 credits/ $0.26, up to 6 min). Stable Audio 2.0 is deprecated. Known for high-quality instrumental generation with precise duration control.

3D 2 models

2 ungraded

3D pricing page ↗

Stability AI's Developer Platform (platform.stability.ai). Credit-metered: 1 credit = $0.01 USD; 25 free credits on signup. Two single-image-to-3D endpoints: Stable Fast 3D (10 credits, ~0.5s) and Stable Point Aware 3D / SPAR3D (4 credits, adds point-cloud sampling for back-side structure). Both models are open-weights (Community License; Enterprise License above $1M revenue) and self-hostable — the platform API is the metered hosted option. No texture/rig/remesh endpoints; the 3D models output a UV-unwrapped textured GLB directly.

See how Stability AI's models fit into a cost-aware routing setup

Release cadence

7 models tracked across all verticals. Earliest release 2024-04, most recent 2025-01.

Derived from Modelglass's per-vertical provider registries — image/llm/video/audio's registry/providers/*.yaml records, plus science/coding/agentic/gdpval's benchmark registries, where a provider is identified by the prefix before "/" in each model's model_id rather than a registry record of its own. Where the same company uses a different provider slug or model_id prefix across verticals (e.g. Google DeepMind vs. Google Cloud, or Qwen's "alibaba" model_id prefix in the coding registry vs. its "qwen" provider slug elsewhere), coverage is merged onto a single profile per an explicit, human-reviewed alias list (registry/provider-aliases.yaml) — everything else reflects exact-slug or exact-prefix matches only.