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 Agnes-2.5-Pro-Theta

120 open-weight models that do the same job as Agnes-2.5-Pro-Theta (188B 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
Agnes-AI logo

Agnes-2.5-Pro-Theta

Agnes-AI·Vision

≈117 GB

Agnes 2.5 Pro Theta is a multimodal mixture-of-experts checkpoint distributed with its model implementation, tokenizer, chat template, and image/video processors. This release provides BF16 weights for local inference through Transformers.

188BApache 2.0
813Nov 2026
--vram
any
--gpus
--quant
  1. nvidia logo≈1.4 GB

    NVIDIA Nemotron Parse 2.0 transforms document images into structured, machine-readable representations with text, layout classes, bounding boxes, and reading-order information.

    903M9K ctx Custom licence
    525.4KJun 2026
    • Same category
    • Smaller (903M vs 188B)
    • Commercial use OK
    vs
  2. internlm logo

    Intern-S2-Mobius

    internlm·Vision

    ≈22 GB

    We introduce Intern-S2-Mobius, a 35B foundation model built on the Mobius-v0 architecture realized by Xtuner and LMDeploy.

    36B256K ctx Apache 2.0
    1.1KJul 2026
    • Same category
    • Smaller (36B vs 188B)
    • Commercial use OK
    vs
  3. internlm logo

    Intern-S2-Preview

    internlm·Vision

    ≈23 GB

    We introduce Intern-S2-Preview, an efficient 35B scientific multimodal foundation model. Beyond conventional parameter and data scaling, Intern-S2-Preview explores task scaling: increasing the difficulty, diversity, and coverage of scientific tasks to further unlock model capabilities.

    36.1BA4.9B256K ctx Apache 2.0
    1.7KMay 2026
    • Same category
    • Smaller (36.1B vs 188B)
    • Commercial use OK
    vs
  4. orcarouter logo

    Qwen3.8-27B-Uncensored

    orcarouter·Vision

    ≈19 GB

    One Gateway. Every Model. — Route Smarter · Ship Safer · Spend Less.

    27.8BApache 2.0
    8.3KAug 2026gguf
    • Same category
    • Smaller (27.8B vs 188B)
    • Commercial use OK
    vs
    • Same category
    • Smaller (40B vs 188B)
    • Commercial use OK
    vs
  5. 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
    3MOct 2025ollama
    • Same category
    • Smaller (4.4B vs 188B)
    • Commercial use OK
    • Widely used
    vs
  6. nvidia logo

    Flora-Rewarder

    nvidia·Vision

    ≈19 GB

    Flora-rewarder-27B produces outcome rewards (an episode-level success score), used for monitoring real-world deployments, and process rewards (dense action-quality scores over short time steps), used during reinforcement learning-based post-training.

    27.8B256K ctx NVIDIA Non-Commercial License
    5.5KOct 2026
    • Same category
    • Smaller (27.8B vs 188B)
    vs
  7. Qwen logo

    Qwen3.5-0.8B-Base

    Qwen·Vision

    ≈1.4 GB

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

    873M256K ctx Apache 2.0
    872.1KFeb 2026gguf
    • Same category
    • Smaller (873M vs 188B)
    • Commercial use OK
    vs
  8. nvidia logo

    Flora-Captioner

    nvidia·Vision

    ≈19 GB

    Flora-captioner-27B generates one action instruction for a supplied target interval in a robot or human manipulation video. It uses neighboring video context to disambiguate the intended local operation.

    27.4B256K ctx NVIDIA Non-Commercial License
    5.6KOct 2026
    • Same category
    • Smaller (27.4B vs 188B)
    vs
  9. naver-hyperclovax logo

    HyperCLOVAX-SEED-Think-32B

    naver-hyperclovax·Vision

    ≈23 GB

    HyperCLOVA X SEED 32B Think is an updated vision-language thinking model that advances the SEED Think 14B line beyond simple scaling, pairing a unified vision-language Transformer backbone with a reasoning-centric training recipe.

    33.3B128K ctx hyperclovax
    7KDec 2025gguf
    • Same category
    • Smaller (33.3B vs 188B)
    vs
  10. nvidia logo≈22 GB

    NVIDIA-Ising-Calibration-1 analyzes quantum computing calibration experiment plots and generates structured technical text across six analysis question categories. NVIDIA-Ising-Calibration-1 was developed by NVIDIA as a quantum calibration vision-language model built on Qwen3.5-35B-A3B.

    35BA3B256K ctx NVIDIA Open Model License
    1.9KMar 2026gguf
    • Same category
    • Smaller (35B vs 188B)
    • Commercial use OK
    vs
  11. BAAI logo

    AREX-2

    BAAI·Vision

    ≈19 GB

    AREX-2 is a 27B-parameter long-horizon agent model from the Beijing Academy of Artificial Intelligence (BAAI). It learns to improve a solution over multiple test-time rounds: propose, measure, reflect, and revise.

    27.4B256K ctx Apache 2.0
    3.8KSept 2026gguf
    • Same category
    • Smaller (27.4B vs 188B)
    • Commercial use OK
    vs
  12. OpenGVLab logo

    InternVL3_5-38B

    OpenGVLab·Vision

    ≈25 GB

    We introduce InternVL3.5, a new family of open-source multimodal models that significantly advances versatility, reasoning capability, and inference efficiency along the InternVL series.

    38.4B40K ctx Apache 2.0
    3.7KAug 2025gguf
    • Same category
    • Smaller (38.4B vs 188B)
    • Commercial use OK
    vs
  13. google logo

    gemma-3-27b-it

    google·Vision

    ≈21 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.

    27.4B128K ctx Gemma Terms of Use
    291KMar 2025ollama
    • Same category
    • Smaller (27.4B vs 188B)
    vs
  14. meta-llama logo

    Llama-3.2-90B-Vision

    meta-llama·Vision

    ≈56 GB

    The Llama 3.2-Vision collection of multimodal large language models (LLMs) is a collection of pretrained and instruction-tuned image reasoning generative models in 11B and 90B sizes (text \+ images in / text out).

    88.6B128K ctx Llama 3.2 Community License
    22Sept 2024
    • Same category
    • Smaller (88.6B vs 188B)
    vs
  15. 🔥 Update (April 5): I’ve released the complete training notebook, codebase, and a comprehensive PDF guide to help beginners and enthusiasts understand and reproduce this model's fine-tuning process.

    27.8B256K ctx Apache 2.0
    14.2KFeb 2026gguf
    • Same category
    • Smaller (27.8B vs 188B)
    • Commercial use OK
    vs
  16. 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
    504KFeb 2026gguf
    • Same category
    • Smaller (4.7B vs 188B)
    • Commercial use OK
    vs
  17. Zyphra logo

    ZAYA1-VL-8B

    Zyphra·Vision

    ≈6.6 GB

    ZAYA1-VL-8B is a vision-language model (VLM) built upon Zyphra's ZAYA1-8B LLM. It has state-of-the-art performance among VLMs for its size and inference efficiency.

    9.7B32K ctx Apache 2.0
    337.9KApr 2026gguf
    • Same category
    • Smaller (9.7B vs 188B)
    • Commercial use OK
    vs
  18. meta-models logo

    Muse-Glimmer-30B-assistant

    meta-models·Vision

    ≈2.2 GB

    Muse Glimmer is a 30-billion-parameter causal language model with a dedicated perception encoder, distilled from Muse Spark and purpose-built for autonomous agentic tasks on consumer hardware.

    2.6B128K ctx Apache 2.0
    129.6KAug 2026
    • Same category
    • Smaller (2.6B vs 188B)
    • Commercial use OK
    vs
  19. Qwen logo≈2.6 GB

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

    2.1B256K ctx Apache 2.0
    1.1MOct 2025ollama
    • Same category
    • Smaller (2.1B vs 188B)
    • Commercial use OK
    vs
  20. fastino logo

    GLiDE

    fastino·Vision

    ≈18 GB

    GLiDE is a decision model from Fastino. Given a state (text or JSON, optionally with images) and one or more typed questions, it returns a probability for every option of every question. Internal use only.

    26.1B256K ctx fastino-community
    3.6KOct 2026
    • Same category
    • Smaller (26.1B vs 188B)
    vs
  21. h2oai logo

    h2o-lightning-31b

    h2oai·Vision

    ≈21 GB

    H2O-Lightning-31B is a 31B-parameter decision model from H2O.ai, built on google/gemma-4-31B-it. It answers typed decision questions about a record (a document, a ticket, a policy, a conversation), and about images that come with the record. It returns a probability for every option:

    31.3Bh2o-model-non-commercial-licence-v1.1
    690Oct 2026
    • Same category
    • Smaller (31.3B vs 188B)
    vs
  22. fastino logo

    GLiDE-27B

    fastino·Vision

    ≈18 GB

    Inference provider: Fastino — use GLiDE through Fastino's hosted API (POST https://api.fastino.ai/v1/systemone, model fastino/GLiDE). Get an API key · API documentation

    26.1B256K ctx fastino-community
    3.6KOct 2026gguf
    • Same category
    • Smaller (26.1B vs 188B)
    vs
  23. Agnes-AI logo

    Agnes-3.0-Flash

    Agnes-AI·Vision

    ≈22 GB

    This repository contains an earlier open-weight Preview checkpoint of Agnes 3.0 Flash. It is distinct from the newer production/API checkpoint listed on Artificial Analysis. The Preview release has 33B parameters and a context window of 262,144 tokens.

    33.1B256K ctx Apache 2.0
    485Sept 2026gguf
    • Same category
    • Smaller (33.1B vs 188B)
    • Commercial use OK
    vs