AI Model Timeline

Tracking the accelerating release frequency of frontier AI models.

Stable Diffusion 3.5 Medium

Stability AI·

ImageimagelocalMobile / NPU

Overview

Compact 2.5B parameter variant optimized for consumer GPUs. Runs on hardware with 4-16 GB VRAM.

Capabilities and innovations

2.5B parametersText-to-imageConsumer GPU optimized (4-16 GB VRAM)Open weights (Stability Community License)Efficient MMDiT at 2.5B ScaleConsumer Hardware Target

Benchmarks

No benchmark values are recorded for this model.

  • GPQA Diamond: not reported by the vendor
  • Humanity’s Last Exam (no tools): not reported by the vendor
  • SWE-bench Verified: not reported by the vendor
  • SWE-bench Pro: benchmark did not exist at release
  • MMLU: not reported by the vendor
  • MMLU-Pro: not reported by the vendor
  • Terminal-Bench 2.x: benchmark did not exist at release
  • MMMU-Pro: benchmark did not exist at release

Architecture and hardware

Parameters
2.5B total (Dense)
Estimated VRAM at Q4
~2 GB, Mobile / NPU class

Links

More from Stability AI

Data curated by AI Model Timeline. See the methodology for admission criteria, benchmark eras and source priority.