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 DeskForge-Gemma4-E4B

120 open-weight models that do the same job as DeskForge-Gemma4-E4B (7.9B 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.

the original
docling-project logo

DeskForge-Gemma4-E4B

docling-project·Vision

≈5.9 GB

DeskForge-Gemma4-E4B is Gemma 4 E4B (instruction-tuned) fine-tuned on 200K grounding examples from DeskForge-1M. Given a desktop screenshot and a natural-language instruction, it answers with the next action as a pyautogui call whose coordinates are fractions of the screen.

7.9B128K ctx Apache 2.0
449Oct 2026gguf
--vram
any
--gpus
--quant
  1. ATH-MaaS logo

    Ovis2.5-9B

    ATH-MaaS·Vision

    ≈7.1 GB

    We are pleased to announce the release of Ovis2.5, the successor to Ovis2, designed for native-resolution visual perception and enhanced multimodal reasoning.

    9.2B40K ctx Apache 2.0
    4.9KAug 2025
    • Same category
    • Similar size (9.2B vs 7.9B)
    • Commercial use OK
    vs
  2. 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 7.9B)
    • Commercial use OK
    vs
  3. HuggingFaceM4 logo

    Idefics3-8B-Llama3

    HuggingFaceM4·Vision

    ≈6.5 GB

    Idefics3 is an open multimodal model that accepts arbitrary sequences of image and text inputs and produces text outputs. The model can answer questions about images, describe visual content, create stories grounded on multiple images, or simply behave as a pure language model without visual…

    8.5B128K ctx Apache 2.0
    37.9KAug 2024gguf
    • Same category
    • Similar size (8.5B vs 7.9B)
    • Commercial use OK
    vs
  4. OpenMOSS-Team logo

    MOSS-VL-Realtime

    OpenMOSS-Team·Vision

    ≈8.7 GB

    MOSS-VL is an open vision-language model family from OpenMOSS, supporting image understanding, long-video understanding, and realtime streaming interaction.

    11.3B256K ctx Apache 2.0
    2.2KJul 2026
    • Same category
    • Similar size (11.3B vs 7.9B)
    • Commercial use OK
    vs
  5. ByteDance logo

    EvoQuality

    ByteDance·Vision

    ≈5.8 GB

    EvoQuality is an open-weight vision model with 8.3B parameters from ByteDance, published on Hugging Face.

    8.3B125K ctx Apache 2.0
    149Jun 2026gguf
    • Same category
    • Similar size (8.3B vs 7.9B)
    • Commercial use OK
    vs
  6. FreedomIntelligence logo

    HuatuoGPT-Vision-7B-Qwen2.5VL

    FreedomIntelligence·Vision

    ≈5.8 GB

    This is the HuatuoGPT-Vision model based on the Qwen2.5-VL architecture. It was trained on the PubMedVision dataset using the Qwen2.5-VL framework. For details, please refer to the training code at HuatuoGPT-Vision Code.

    8.3B125K ctx Apache 2.0
    3KMay 2025gguf
    • Same category
    • Similar size (8.3B vs 7.9B)
    • Commercial use OK
    vs
  7. lmms-lab logo≈5.7 GB

    The LLaVA-OneVision models are 0.5/7/72B parameter models trained on LLaVA-OneVision, based on Qwen2 language model with a context window of 32K tokens.

    8B32K ctx Apache 2.0
    28.7KJun 2024
    • Same category
    • Similar size (8B vs 7.9B)
    • Commercial use OK
    vs
  8. OpenGVLab logo

    InternVL3-9B

    OpenGVLab·Vision

    ≈6.3 GB

    We introduce InternVL3, an advanced multimodal large language model (MLLM) series that demonstrates superior overall performance.

    9.1B32K ctx MIT
    10.2KApr 2025
    • Same category
    • Similar size (9.1B vs 7.9B)
    • Commercial use OK
    vs
  9. Qwen logo

    Qwen3.8-27B

    Qwen·Vision

    ≈19 GB

    Following the widespread community adoption of the Qwen3.5 and Qwen3.6 series, we are pleased to introduce Qwen3.8, the most capable generation in the Qwen open-model family to date.

    27.8B256K ctx Apache 2.0
    6.3MAug 2026ollama
    • Same category
    • Larger (27.8B vs 7.9B)
    • Commercial use OK
    • Widely used
    vs
  10. 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 7.9B)
    vs
  11. allenai logo

    Molmo-7B-D-0924

    allenai·Vision

    ≈5.5 GB

    Molmo is a family of open vision-language models developed by the Allen Institute for AI. Molmo models are trained on PixMo, a dataset of 1 million, highly-curated image-text pairs. It has state-of-the-art performance among multimodal models with a similar size while being fully open-source.

    8B4K ctx Apache 2.0
    11.9KSept 2024
    • Same category
    • Similar size (8B vs 7.9B)
    • Commercial use OK
    vs
  12. 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 7.9B)
    vs
  13. 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 7.9B)
    • Commercial use OK
    vs
  14. 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 7.9B)
    • Commercial use OK
    vs
  15. inclusionAI logo

    ZwZ-8B

    inclusionAI·Vision

    ≈6.8 GB

    ZwZ-8B is a fine-grained multimodal perception model built upon Qwen3-VL-8B. It is trained using Region-to-Image Distillation (R2I) combined with reinforcement learning, enabling superior fine-grained visual understanding in a single forward pass — no inference-time zooming or tool calling required.

    8.8B256K ctx Apache 2.0
    517Feb 2026gguf
    • Same category
    • Similar size (8.8B vs 7.9B)
    • Commercial use OK
    vs
  16. 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 7.9B)
    • Commercial use OK
    vs
  17. openbmb logo

    MiniCPM-V-2_6

    openbmb·Vision

    ≈6.3 GB

    A GPT-4V Level MLLM for Single Image, Multi Image and Video on Your Phone

    8.1B32K ctx Unknown licence
    17KAug 2024gguf
    • Same category
    • Similar size (8.1B vs 7.9B)
    vs
  18. BAAI logo

    Video-XL-2

    BAAI·Vision

    ≈5.8 GB

    TODO - [X] Release model weights. - [X] Release the inference code w/o. efficiency optimization. - [X] Release the inference code w. chunk-based prefill. - [ ] Release the inference code w. chunk-based prefill & bi-level kvs decoding.

    8.1B32K ctx Apache 2.0
    1.7KMay 2025
    • Same category
    • Similar size (8.1B vs 7.9B)
    • Commercial use OK
    vs
  19. lmms-lab logo

    LLaVA-Video-7B-Qwen2

    lmms-lab·Vision

    ≈5.7 GB

    The LLaVA-Video models are 7/72B parameter models trained on LLaVA-Video-178K and LLaVA-OneVision Dataset, based on Qwen2 language model with a context window of 32K tokens.

    8B32K ctx Apache 2.0
    12.3KSept 2024
    • Same category
    • Similar size (8B vs 7.9B)
    • Commercial use OK
    vs
  20. apple logo

    FastVLM-7B

    apple·Vision

    ≈5.5 GB

    FastVLM was introduced in FastVLM: Efficient Vision Encoding for Vision Language Models. (CVPR 2025)

    7.8B32K ctx Apple Machine Learning Research License
    775Aug 2025
    • Same category
    • Similar size (7.8B vs 7.9B)
    vs
  21. jdopensource logo

    JoyAI-VL-Interaction-Preview

    jdopensource·Vision

    ≈6.8 GB

    The first open, vision-driven real-time interaction model — it watches a live video stream and decides on its own when to speak, stay silent, or delegate.

    8.8B256K ctx Apache 2.0
    160Jun 2026gguf
    • Same category
    • Similar size (8.8B vs 7.9B)
    • Commercial use OK
    vs
  22. 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 7.9B)
    • Commercial use OK
    vs
  23. tencent logo

    Penguin-VL-8B

    tencent·Vision

    ≈6.8 GB

    Penguin-VL Exploring the Efficiency Limits of VLM with LLM-based Vision Encoders

    8.7B40K ctx Apache 2.0
    307Mar 2026
    • Same category
    • Similar size (8.7B vs 7.9B)
    • Commercial use OK
    vs
  24. XiaomiMiMo logo

    MiMo-VL-7B-RL

    XiaomiMiMo·Vision

    ≈6.5 GB

    In this report, we share our efforts to build a compact yet powerful VLM, MiMo-VL-7B. MiMo-VL-7B comprises (1) a native resolution ViT encoder that preserves fine-grained visual details, (2) an MLP projector for efficient cross-modal alignment, and (3) our MiMo-7B language model, specifically…

    8.3B125K ctx MIT
    1.8KMay 2025gguf
    • Same category
    • Similar size (8.3B vs 7.9B)
    • Commercial use OK
    vs