Stable Diffusion 3.5
The open-weights image workhorse: 8B MMDiT, great prompt adherence.
About
Stable Diffusion 3.5 Large (stabilityai/stable-diffusion-3.5-large), released October 2024, is Stability AI's 8.1B-parameter MMDiT image model — a big leap in prompt adherence and typography over SDXL, runnable on consumer GPUs via Diffusers/ComfyUI. Weights are downloadable (gated; name/email required) under the Stability AI Community License — custom, non-OSI: free for research and commercial use within its terms, but NOT open source. Describe as "open weights". FLUX and newer generations have since taken the quality crown, but SD 3.5 remains the most fine-tunable, ecosystem-rich open-weights image model.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
No standard benchmarks; 2024 evaluations showed strong prompt adherence vs SDXL/FLUX.1-dev tradeoffs (quality near dev, faster) — qualitative, test your own prompts.
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 7/10
Scored 7/10 (verified tier, metrics-v1). Evidence: HF downloads 47,695 (peer-normalized); released 721 days ago. Manual review: open-weights image workhorse (8B MMDiT, great prompt adherence); superseded by newer FLUX/SD generations. License: Stability AI Community License — custom, non-OSI (gated, name/email required); describe as 'open weights', never 'open source'.
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 7.1/10
- Maintenance · recency, releases, issue load
- 5.0/10
- Openness · license + docs
- 6.5/10
- Manual review · sanity read + red-flag check
- 8.0/10
- · HF downloads 47,695 (peer-normalized)
- · released 721 days ago
- · manual review: open-weights image workhorse (8B MMDiT, great prompt adherence); superseded by newer FLUX/SD generations