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 Axion-1.0-35B

120 open-weight models that do the same job as Axion-1.0-35B (35.1B chat), 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
Axion1-0 logo

Axion-1.0-35B

Axion1-0·Chat

≈22 GB

This repository provides a reproduction checkpoint of Axion 1.0-35B, following the recipe described in Axion 1.0: Open Agent Foundation Models, Simplified. The model is designed for scientific reasoning, information seeking, and tool-augmented problem solving.

35.1BA3.9B256K ctx Apache 2.0
1.8KOct 2026gguf
--vram
any
--gpus
--quant
  1. microsoft logo

    Phi-3.5-MoE-instruct

    microsoft·Chat

    ≈26 GB

    Phi-3.5-MoE is a lightweight, state-of-the-art open model built upon datasets used for Phi-3 - synthetic data and filtered publicly available documents - with a focus on very high-quality, reasoning dense data. The model supports multilingual and comes with 128K context length (in tokens).

    41.9BA6.6B128K ctx MIT
    20.8KAug 2024gguf
    • Same category
    • Similar size (41.9B vs 35.1B)
    • Commercial use OK
    vs
  2. NousResearch logo

    nomos-1

    NousResearch·Chat

    ≈19 GB

    We release Nomos 1, a specialization of Qwen/Qwen3-30B-A3B-Thinking-2507 for mathematical problem-solving and proof-writing in natural language. Nomos-1 was trained in collaboration with Hillclimb AI.

    30.5BA3.4B256K ctx Apache 2.0
    409Dec 2025gguf
    • Same category
    • Similar size (30.5B vs 35.1B)
    • Commercial use OK
    vs
  3. 01-ai logo

    Yi-1.5-34B-32K

    01-ai·Chat

    ≈23 GB

    Yi-1.5 is an upgraded version of Yi. It is continuously pre-trained on Yi with a high-quality corpus of 500B tokens and fine-tuned on 3M diverse fine-tuning samples.

    34.4B32K ctx Apache 2.0
    9.7KMay 2024gguf
    • Same category
    • Similar size (34.4B vs 35.1B)
    • Commercial use OK
    vs
  4. 01-ai logo≈23 GB

    Yi-1.5 is an upgraded version of Yi. It is continuously pre-trained on Yi with a high-quality corpus of 500B tokens and fine-tuned on 3M diverse fine-tuning samples.

    34.4B16K ctx Apache 2.0
    9.6KMay 2024gguf
    • Same category
    • Similar size (34.4B vs 35.1B)
    • Commercial use OK
    vs
  5. Alibaba-NLP logo

    WebDancer-32B

    Alibaba-NLP·Chat

    ≈22 GB

    This model was presented in the paper WebDancer: Towards Autonomous Information Seeking Agency.

    32.8B40K ctx MIT
    701Jun 2025gguf
    • Same category
    • Similar size (32.8B vs 35.1B)
    • Commercial use OK
    vs
  6. zai-org logo

    GLM-4-32B-Base-0414

    zai-org·Chat

    ≈20 GB

    The GLM family welcomes new members, the GLM-4-32B-0414 series models, featuring 32 billion parameters. Its performance is comparable to OpenAI’s GPT series and DeepSeek’s V3/R1 series. It also supports very user-friendly local deployment features.

    32.6B32K ctx MIT
    1.1KApr 2025gguf
    • Same category
    • Similar size (32.6B vs 35.1B)
    • Commercial use OK
    vs
  7. 01-ai logo

    Yi-34B-Chat

    01-ai·Chat

    ≈22 GB

    The Yi series models adopt the same model architecture as Llama but are NOT derivatives of Llama.

    34.4B4K ctx Apache 2.0
    28.5KNov 2023ollama
    • Same category
    • Similar size (34.4B vs 35.1B)
    • Commercial use OK
    vs
  8. deepgrove logo

    maple-preview

    deepgrove·Chat

    ≈13 GB

    Today we introduce Maple-Preview, an open-source 20B-A1B ternary-weight reasoning LLM. Maple-Preview has SOTA reasoning for its weight class and is even competitive with larger models.

    20.2BA1.5B128K ctx MIT
    2KAug 2026gguf
    • Same category
    • Similar size (20.2B vs 35.1B)
    • Commercial use OK
    vs
  9. CohereLabs logo

    aya-expanse-32b

    CohereLabs·Chat

    ≈22 GB

    Aya Expanse 32B is an open-weight research release of a model with highly advanced multilingual capabilities. It focuses on pairing a highly performant pre-trained Command family of models with the result of a year’s dedicated research from Cohere Labs, including data arbitrage, multilingual…

    32.3BCC BY-NC 4.0
    3.5KOct 2024ollama
    • Same category
    • Similar size (32.3B vs 35.1B)
    vs
  10. baidu logo≈14 GB

    The advanced capabilities of the ERNIE 4.5 models, particularly the MoE-based A47B and A3B series, are underpinned by several key technical innovations:

    21.9BA3B128K ctx Apache 2.0
    16.6KJun 2025gguf
    • Same category
    • Similar size (21.9B vs 35.1B)
    • Commercial use OK
    vs
  11. perplexity-ai logo

    browsesafe

    perplexity-ai·Chat

    ≈19 GB

    BrowseSafe is a multi-layered defense strategy comprising both architectural and model-based defenses to protect against evolving prompt injection attacks. It is a specialized security model designed to protect AI browser agents from prompt injection attacks embedded in real-world web content.

    30.5BA3.4B256K ctx MIT
    277Nov 2025gguf
    • Same category
    • Similar size (30.5B vs 35.1B)
    • Commercial use OK
    vs
  12. utter-project logo

    EuroLLM-22B-2512

    utter-project·Chat

    ≈16 GB

    This is the model card for EuroLLM-22B. You can also check the post-trained version: EuroLLM-22B-Instruct-2515.

    22.6B32K ctx Apache 2.0
    4.1KNov 2025gguf
    • Same category
    • Similar size (22.6B vs 35.1B)
    • Commercial use OK
    vs
  13. Dolphin Mistral 24B Venice Edition is a collaborative project we undertook with Venice.ai with the goal of creating the most uncensored version of Mistral 24B for use within the Venice ecosystem.

    24B128K ctx Apache 2.0
    7.4KJun 2025gguf
    • Same category
    • Similar size (24B vs 35.1B)
    • Commercial use OK
    vs
  14. inclusionAI logo

    GroveMoE-Inst

    inclusionAI·Chat

    ≈21 GB

    🤗 Models &nbsp&nbsp | &nbsp&nbsp 📑 Paper &nbsp&nbsp | &nbsp&nbsp 🔗 Github &nbsp&nbsp

    32.9BA5.8B40K ctx Apache 2.0
    248Aug 2025gguf
    • Same category
    • Similar size (32.9B vs 35.1B)
    • Commercial use OK
    vs
  15. moonshotai logo

    Kimi-Linear-48B-A3B-Base

    moonshotai·Chat

    ≈32 GB

    (a) On MMLU-Pro (4k context length), Kimi Linear achieves 51.0 performance with similar speed as full attention. On RULER (128k context length), it shows Pareto-optimal performance (84.3) and 3.98x speedup.

    49.1BA3B1M ctx MIT
    1.8KOct 2025gguf
    • Same category
    • Similar size (49.1B vs 35.1B)
    • Commercial use OK
    vs
  16. deepcogito logo≈22 GB

    The Cogito LLMs are instruction tuned generative models (text in/text out). All models are released under an open license for commercial use.

    32.8B128K ctx Apache 2.0
    709Mar 2025gguf
    • Same category
    • Similar size (32.8B vs 35.1B)
    • Commercial use OK
    vs
  17. ai21labs logo

    AI21-Jamba2-Mini

    ai21labs·Chat

    ≈32 GB

    Jamba2 Mini is an open source small language model built for enterprise reliability. With 12B active parameters (52B total), it delivers precise question answering without the computational overhead of reasoning models.

    51.6B256K ctx Apache 2.0
    1.3KJan 2026gguf
    • Same category
    • Similar size (51.6B vs 35.1B)
    • Commercial use OK
    vs
  18. Salesforce logo

    xLAM-2-32b-fc-r

    Salesforce·Chat

    ≈22 GB

    Large Action Models (LAMs) are advanced language models designed to enhance decision-making by translating user intentions into executable actions.

    32.8B32K ctx CC BY-NC 4.0
    654Mar 2025gguf
    • Same category
    • Similar size (32.8B vs 35.1B)
    vs
  19. inclusionAI logo

    LLaDA2.1-mini

    inclusionAI·Chat

    ≈10 GB

    🚀 LLaDA2.1-flash is now live on ZenmuxAI! Try it via API 🛠️ or Chat 💬: https://zenmux.ai/inclusionai/llada2.1-flash

    16.3BA1.4B32K ctx Apache 2.0
    28.2KFeb 2026
    • Same category
    • Smaller (16.3B vs 35.1B)
    • Commercial use OK
    vs
  20. poolside logo

    Laguna-S-2.1

    poolside·Chat

    ≈72 GB

    Laguna S 2.1 is a 118B total parameter Mixture-of-Experts model with 8B activated parameters per token, designed for agentic coding and long-horizon work.

    118BA6.1B1M ctx Custom licence
    488.2KJul 2026ollama
    • Same category
    • Larger (118B vs 35.1B)
    • Commercial use OK
    vs
  21. google logo

    gemma-2-27b-it

    google·Chat

    ≈20 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.

    27.2B8K ctx Gemma Terms of Use
    23.2KJun 2024ollama
    • Same category
    • Similar size (27.2B vs 35.1B)
    vs
  22. ornith-ai logo

    Ornith-1.5-122B

    ornith-ai·Chat

    ≈75 GB

    Chirp Chirp! 🐦 Meet Ornith-1.5-122B: No Tradeoff Between Strong and Fast Stronger. Faster. Leaner. It delivers stronger benchmark results, finishes tasks faster, and produces shorter outputs, bringing better instruction following, stronger task decomposition, and deeper knowledge of real-world…

    125BA12.7B256K ctx MIT
    341.7KOct 2026ollama
    • Same category
    • Larger (125B vs 35.1B)
    • Commercial use OK
    vs
  23. BSC-LT logo

    ALIA-40b

    BSC-LT·Chat

    ≈26 GB

    ALIA-40b is a highly multilingual model pre-trained from scratch that will come with its respective base and instruction-tuned variants. This model card corresponds to the 40b base version.

    40.4B32K ctx Apache 2.0
    2.1KDec 2024gguf
    • Same category
    • Similar size (40.4B vs 35.1B)
    • Commercial use OK
    vs
  24. meta-llama logo

    Llama-3.3-70B-Instruct

    meta-llama·Chat

    ≈45 GB

    The Meta Llama 3.3 multilingual large language model (LLM) is an instruction tuned generative model in 70B (text in/text out).

    70.6B128K ctx Llama 3.3 Community License
    344.9KNov 2024ollama
    • Same category
    • Larger (70.6B vs 35.1B)
    vs