USASI
Model family

FastVLM

Family overview — summarizes releases; licenses and availability belong to each release.

Maintained by Apple354

FastVLM is a family of vision-language models from Apple, described in a CVPR 2025 paper, built around FastViTHD, a hybrid convolutional-transformer vision encoder that outputs fewer visual tokens to reduce encoding time for high-resolution images. Apple released 0.5B, 1.5B, and 7B variants that pair the encoder with Qwen2 language models, with PyTorch checkpoints and versions exported for Apple silicon.145

Last reviewedEntry updated Documented release Unknown

Releases assessed

Each release is assessed on its own. This family has no family-wide openness label.
Releases in the FastVLM family
ReleaseWeightsTier (USASI rubric v0.1)LicenseReleased
FastVLM 7BWeights: PublicModel-disclosure tier (USASI rubric v0.1): Open-weightApple Machine Learning Research Model License Agreement, Apple software license (ml-fastvlm code)Unknown

What it is useful for

Apple describes the models for image understanding with low time-to-first-token and for on-device use; the repository includes an iOS and macOS demo app.51

Organization context

U.S. eligibility

Project eligibility rests on documented governing or maintaining entities, not on contributors.

Eligible · basis: U.S. headquarters

The model license states the models are developed and released by Apple Inc., and the paper lists Apple as the authors' affiliation. Apple Inc. has its principal executive offices in Cupertino, California, per its Form 10-K for fiscal year 2025. The language model components are third-party Qwen2 models from the Qwen Team at Alibaba Group, recorded on the release record; they are not U.S.-developed.3476

Assessed Sep 29, 2026

Sources

  1. 1.
    apple/ml-fastvlm README (external site: raw.githubusercontent.com)

    Apple (GitHub) · Repository · accessed Sep 29, 2026

  2. 2.
    apple/ml-fastvlm LICENSE (code) (external site: raw.githubusercontent.com)

    Apple Inc. (GitHub) · License · accessed Sep 29, 2026

  3. 3.
  4. 4.
  5. 5.
    FastVLM: Efficient Vision Encoding for Vision Language Models (external site: machinelearning.apple.com)

    Apple Machine Learning Research · Announcement · published Jul 23, 2025 · accessed Sep 29, 2026

  6. 6.
    Qwen2 Technical Report (arXiv 2407.10671, HTML) (external site: arxiv.org)

    arXiv · Paper · published Jul 15, 2024 · accessed Sep 29, 2026

  7. 7.
    Apple Inc. Form 10-K for the fiscal year ended September 27, 2025 (external site: sec.gov)

    Apple Inc. (via U.S. Securities and Exchange Commission EDGAR) · Filing · accessed Sep 29, 2026

This listing is not an endorsement, a safety assessment, or a federal approval.

Support Us

Help keep USASI useful.

Optional. No USASI account required. Payment takes place on the linked provider’s website (Buy Me a Coffee).

About supporting this project