selfhost.directory

The index of self-hostable projects — ranked, categorized and served with live updates.
Own your data and stop giving it to big tech.

Alternatives to ZwZ2-4B

120 open-weight models that do the same job as ZwZ2-4B (4.5B vision), ranked by how close they are in task and size, then by popularity and how recent they are. Licence and VRAM are shown for each.

Compare ZwZ2-4B with the top 3
the original
inclusionAI logo

ZwZ2-4B

inclusionAI·Vision

≈4.2 GB

ZwZ2-4B is the Qwen3.5-4B-based model in Zooming-without-Zooming-v2 (ZwZ2). It learns to connect visual evidence across regions through comparison, aggregation, matching, and multi-step reasoning. Training uses full images; inference requires no zooming tools.

4.5B256K ctx Unknown licence
1.1KOct 2026gguf
--vram
any
--gpus
--quant
  1. Qwen logo

    Qwen3.5-4B

    Qwen·Vision

    ≈4.3 GB

    Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.

    4.7B256K ctx Apache 2.0
    4.3MFeb 2026ollama
    • Same category
    • Similar size (4.7B vs 4.5B)
    • Commercial use OK
    • Widely used
    vs
  2. Qwen logo≈4.3 GB

    Meet Qwen3-VL — the most powerful vision-language model in the Qwen series to date.

    4.4B256K ctx Apache 2.0
    2.8MOct 2025ollama
    • Same category
    • Similar size (4.4B vs 4.5B)
    • Commercial use OK
    • Widely used
    vs
  3. Qwen logo

    Qwen3.5-4B-Base

    Qwen·Vision

    ≈4.3 GB

    Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.

    4.7B256K ctx Apache 2.0
    496.7KFeb 2026gguf
    • Same category
    • Similar size (4.7B vs 4.5B)
    • Commercial use OK
    vs
  4. h2oai logo

    h2o-lightning-4b

    h2oai·Vision

    ≈4.2 GB

    H2O-Lightning-4B is a 4B-parameter decision model from H2O.ai, built on Qwen/Qwen3.5-4B. It answers typed decision questions about a record (a document, a ticket, a policy, a conversation) and about images that come with the record (photos, screenshots, scanned documents, charts).

    4.5B256K ctx Apache 2.0
    26.1KOct 2026gguf
    • Same category
    • Similar size (4.5B vs 4.5B)
    • Commercial use OK
    vs
  5. tencent logo

    WeVisOne-VL-4B

    tencent·Vision

    ≈4.5 GB

    WeVisOne-VL (WeChat Vision All-in-One) is a unified vision-language model developed by WeChat Vision, Tencent Inc. Built upon Qwen3-VL, it unifies object-level perception (open-vocabulary object detection, referring localization, object counting, object pointing, and OCR) and pixel-level perception…

    4.8B256K ctx Apache 2.0
    36.4KOct 2026gguf
    • Same category
    • Similar size (4.8B vs 4.5B)
    • Commercial use OK
    • Newer release
    vs
  6. google logo

    medgemma-1.5-4b-it

    google·Vision

    ≈4.1 GB

    * License: The use of MedGemma is governed by the Health AI Developer Foundations terms of use. MedGemma has not been evaluated or optimized for multi-turn applications.

    4.3B128K ctx health-ai-developer-foundations
    183.5KJan 2026ollama
    • Same category
    • Similar size (4.3B vs 4.5B)
    vs
  7. google logo

    gemma-3-4b-it

    google·Vision

    ≈3.7 GB

    Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models.

    4.3B128K ctx Gemma Terms of Use
    1.1MFeb 2025ollama
    • Same category
    • Similar size (4.3B vs 4.5B)
    vs
  8. google logo

    medgemma-4b-it

    google·Vision

    ≈4.1 GB

    This section describes the MedGemma model and how to use it.

    4.3B128K ctx health-ai-developer-foundations
    469.7KMay 2025ollama
    • Same category
    • Similar size (4.3B vs 4.5B)
    vs
  9. ibm-granite logo

    granite-vision-4.1-4b

    ibm-granite·Vision

    ≈3.5 GB

    Model Summary: Granite Vision 4.1 4B is a vision-language model (VLM) that delivers frontier-level performance on structured document extraction tasks — chart extraction, table extraction, and semantic key-value pair extraction — in a compact 4B parameter footprint, providing a lightweight…

    4B128K ctx Apache 2.0
    67.8KApr 2026gguf
    • Same category
    • Similar size (4B vs 4.5B)
    • Commercial use OK
    vs
  10. internlm logo

    Intern-Decision-4B

    internlm·Vision

    ≈4.2 GB

    Intern-Decision-4B is a multimodal structured decision model fine-tuned from Qwen3.5-4B. It accepts a shared state, a schema of named questions, and optional images, and returns an answer distribution for every question in one model forward pass.

    4.5B256K ctx Apache 2.0
    5.9KSept 2026gguf
    • Same category
    • Similar size (4.5B vs 4.5B)
    • Commercial use OK
    vs
  11. Qwen logo

    Qwen-Drive-1.0-4B

    Qwen·Vision

    ≈4.0 GB

    Qwen-Drive-1.0 retains the architecture of the pretrained Qwen3.5 vision-language model and integrates 3D perception, visual question answering, and motion planning within a unified framework.

    4.5BApache 2.0
    6.5KAug 2026gguf
    • Same category
    • Similar size (4.5B vs 4.5B)
    • Commercial use OK
    vs
  12. microsoft logo

    Mage-VL-Cowork

    microsoft·Vision

    ≈4.4 GB

    Mage-VL-Cowork A Full Training Recipe to Learn Hybrid Computer Use from Evolved Harnesses

    4.7B256K ctx Apache 2.0
    5.8KOct 2026gguf
    • Same category
    • Similar size (4.7B vs 4.5B)
    • Commercial use OK
    • Newer release
    vs
  13. Qwen logo≈3.0 GB

    In the past five months since Qwen2-VL’s release, numerous developers have built new models on the Qwen2-VL vision-language models, providing us with valuable feedback. During this period, we focused on building more useful vision-language models.

    3.8B125K ctx Qwen Research License
    1.9MJan 2025ollama
    • Same category
    • Similar size (3.8B vs 4.5B)
    • Widely used
    vs
  14. allenai logo

    Molmo2-4B

    allenai·Vision

    ≈4.5 GB

    Molmo2 is a family of open vision-language models developed by the Allen Institute for AI (Ai2) that support image, video and multi-image understanding and grounding.

    4.9B36K ctx Apache 2.0
    46.2KDec 2025gguf
    • Same category
    • Similar size (4.9B vs 4.5B)
    • Commercial use OK
    vs
  15. microsoft logo

    Mage-VL

    microsoft·Vision

    ≈4.4 GB

    Mage-VL is a codec-native, proactive-streaming multimodal foundation model for image and video understanding, whose visual encoder is trained entirely from scratch at a compact 4B scale.

    4.7B256K ctx Apache 2.0
    6.8KJul 2026gguf
    • Same category
    • Similar size (4.7B vs 4.5B)
    • Commercial use OK
    vs
  16. maritaca-ai logo

    S5-nano

    maritaca-ai·Vision

    ≈4.3 GB

    S5-nano is a 4B-parameter model from Maritaca AI. It has two modes:

    4.7B256K ctx Apache 2.0
    2.3KNov 2026gguf
    • Same category
    • Similar size (4.7B vs 4.5B)
    • Commercial use OK
    • Newer release
    vs
  17. Qwen logo

    Qwen3.5-2B

    Qwen·Vision

    ≈2.2 GB

    Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.

    2.3B256K ctx Apache 2.0
    5.4MFeb 2026ollama
    • Same category
    • Similar size (2.3B vs 4.5B)
    • Commercial use OK
    • Widely used
    vs
  18. microsoft logo

    Fara1.5-4B

    microsoft·Vision

    ≈4.2 GB

    Fara1.5-4B is a multimodal computer use agent (CUA) for web browsers, from Microsoft Research AI Frontiers. It observes the browser through screenshots and acts on the user's behalf by emitting structured tool calls — click, type, scroll, visit URL, web search, and so on — to complete tasks…

    4.5B256K ctx MIT
    2.8KJul 2026gguf
    • Same category
    • Similar size (4.5B vs 4.5B)
    • Commercial use OK
    vs
  19. nvidia logo

    LocateAnything-3B

    nvidia·Vision

    ≈3.0 GB

    LocateAnything is a vision-language model for fast and high-quality visual grounding, enabling precise object localization, dense detection, and point-based localization across diverse domains in both Enterprise Intelligence and Physical AI.

    3.8B32K ctx nvidia-license
    31.3KMar 2026gguf
    • Same category
    • Similar size (3.8B vs 4.5B)
    vs
  20. LiquidAI logo

    LFM2.5-VL-3B

    LiquidAI·Vision

    ≈2.8 GB

    LFM2.5-VL-3B is a multimodal variant of LFM2.5, a family of hybrid models designed for on-device deployment. It builds on LFM2-VL-3B with further mid- and post-training.

    3.1B32K ctx lfm1.0
    54.2KAug 2026gguf
    • Same category
    • Similar size (3.1B vs 4.5B)
    vs
  21. Qwen logo

    Qwen3.5-9B

    Qwen·Vision

    ≈7.2 GB

    Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.

    9.7B256K ctx Apache 2.0
    6.2MFeb 2026ollama
    • Same category
    • Larger (9.7B vs 4.5B)
    • Commercial use OK
    • Widely used
    vs
  22. Qwen logo≈6.8 GB

    Meet Qwen3-VL — the most powerful vision-language model in the Qwen series to date.

    8.8B256K ctx Apache 2.0
    6.1MOct 2025ollama
    • Same category
    • Similar size (8.8B vs 4.5B)
    • Commercial use OK
    • Widely used
    vs
  23. google logo

    gemma-3-4b-pt

    google·Vision

    ≈3.7 GB

    Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models.

    4.3BGemma Terms of Use
    116KFeb 2025gguf
    • Same category
    • Similar size (4.3B vs 4.5B)
    vs
  24. nvidia logo

    NV-Reason-CT

    nvidia·Vision

    ≈4.7 GB

    NV-Reason-CT is a 3D vision-language model (VLM) for CT image analysis. It combines a native 3D vision encoder (3D ViT) with a language model and is designed for radiology report generation, general question answering, and multi-step reasoning across chest and abdominal CT volumes.

    5.3B256K ctx Custom licence
    3.9KSept 2026
    • Same category
    • Similar size (5.3B vs 4.5B)
    • Commercial use OK
    vs