Skip to content

EthenEthenEthen

LoginPlatformTry Ethen

Open Source Model Profile · Efficient-Large-Model

Sana_Sprint_1.6B_1024px_diffusers

Sana_Sprint_1.6B_1024px_diffusers is a text-to-image diffusion model from Efficient-Large-Model. According to the model card, it generates 1024px images in 1-4 steps with 1.6B parameters.

Publisher
Efficient-Large-Model
Task
text-to-image
Model type
Unknown
License
apache-2.0
Library
sana, sana-sprint
Publication status
Accepted · not indexed

Model overview

Sana_Sprint_1.6B_1024px_diffusers is published by Efficient-Large-Model as a text-to-image model. According to the model card, developed by NVIDIA and Sana, it is a 1.6B-parameter one-step diffusion model for 1024px-based images with 1-4 step generation.

Recorded capabilities

1-4 step inference design

According to the model card, SANA-Sprint reduces inference steps from 20 to 1-4 with a unified step-adaptive model, training-free sCM distillation, and ControlNet integration.

Publisher-reported speed and scores

According to the model card, the publisher reports 7.59 FID and 0.74 GenEval in 1 step, with 0.1s text-to-image and 0.25s ControlNet latency for 1024x1024 images on H100.

1024px BF16 configuration

According to the model card, the publisher lists 1.6B parameters, torch.bfloat16 precision, and 1024px-based multi-scale image generation.

Diffusers example

According to the model card, the card shows SanaSprintPipeline use with 2 inference steps and states direct use is for research including artworks and design.

Use cases in the source record

  • Fast 1024x1024 text-to-image generation in 1-4 steps using the card's SanaSprintPipeline example.
  • Research generation of artworks and design-process imagery, as stated in the card's direct-use description.

Limitations and unknowns

  • No weight-derived Safetensors parameter count was extracted; the 1.6B figure is a publisher-reported card claim.
  • No independently measured evaluation results were extracted; FID, GenEval, and latency figures are publisher-reported claims.
  • According to the model card, the model was not trained to be factual or true representations of people or events.
  • Provider state is historical snapshot data, not independently refreshed current availability.

Source and provenance

Source: Efficient-Large-Model/Sana_Sprint_1.6B_1024px_diffusers

Captured: Unknown. Processed: 2026-09-07T19:35:04.808891+00:00.

🐱 Sana Model Card Demos Training Pipeline Model Efficiency SANA-Sprint is an ultra-efficient diffusion model for text-to-image (T2I) generation, reducing inference steps from 20 to 1-4 while achieving state-of-the-art performance. Key innovations include: (1) A training-free approach for continuous-time consistency distillation (sCM), eliminating costly retraining; (2) A unified step-adaptive model for high-quality generation in 1-4 steps; and (3) ControlNet integration for real-time interactive image generation. SANA-Sprint achieves 7.59 FID and 0.74 GenEval in just 1 step — outperforming FLUX-schnell (7.94 FID / 0.71 GenEval) whi…

F001F002F003F004F005F006F007F008F009F010F011F012F015F016F017