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 Spark-X2.6-4B-Base

120 open-weight models that do the same job as Spark-X2.6-4B-Base (4.1B 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
XHToken logo

Spark-X2.6-4B-Base

XHToken·Chat

≈4.1 GB

A little over a month has passed since Spark-X2.5. In the rapidly evolving world of AI agents, that is enough time to rethink, refine, and move the frontier forward.

4.1B1M ctx Apache 2.0
2.5KOct 2026gguf
--vram
any
--gpus
--quant
  1. ibm-granite logo≈3.0 GB

    Model Summary: Granite-3.1-3B-A800M-Instruct is a 3B parameter long-context instruct model finetuned from Granite-3.1-3B-A800M-Base using a combination of open source instruction datasets with permissive license and internally collected synthetic datasets tailored for solving long context problems.

    3.3BA800M128K ctx Apache 2.0
    114.9KDec 2024gguf
    • Same category
    • Similar size (3.3B vs 4.1B)
    • Commercial use OK
    vs
  2. LiquidAI logo

    LFM2.5-2.6B-Base

    LiquidAI·Chat

    ≈2.6 GB

    LFM2.5 is a new family of hybrid models designed for on-device deployment. It builds on the LFM2 architecture with extended pre-training and reinforcement learning.

    2.7B125K ctx lfm1.0
    7.4KAug 2026gguf
    • Same category
    • Similar size (2.7B vs 4.1B)
    vs
  3. arcee-ai logo

    AFM-4.5B

    arcee-ai·Chat

    ≈3.8 GB

    AFM-4.5B is a 4.5 billion parameter instruction-tuned model developed by Arcee.ai, designed for enterprise-grade performance across diverse deployment environments from cloud to edge.

    4.6B64K ctx Apache 2.0
    7.1KJul 2025gguf
    • Same category
    • Similar size (4.6B vs 4.1B)
    • Commercial use OK
    vs
  4. ibm-granite logo

    granite-swash-3b-a600m

    ibm-granite·Chat

    ≈2.5 GB

    Model Summary Granite-SWASH-3B-a600M is a 3B parameter decoder-only sparse language model with per-layer Sliding Window Attention (SWA) and learnable attention sinks (LSE-based).

    3BA600M8K ctx Apache 2.0
    3KJul 2026gguf
    • Same category
    • Similar size (3B vs 4.1B)
    • Commercial use OK
    vs
  5. Qwen logo

    Qwen3-1.7B

    Qwen·Chat

    ≈2.6 GB

    Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models.

    2B40K ctx Apache 2.0
    3MApr 2025ollama
    • Same category
    • Smaller (2B vs 4.1B)
    • Commercial use OK
    • Widely used
    vs
  6. janhq logo≈4.2 GB

    Jan-v3-4B-base-instruct is a 4B-parameter model obtained via post-training distillation from a larger teacher, transferring capabilities while preserving general-purpose performance on standard benchmarks.

    4.4B256K ctx Apache 2.0
    1.1KJan 2026gguf
    • Same category
    • Similar size (4.4B vs 4.1B)
    • Commercial use OK
    vs
  7. Qwen logo≈5.5 GB

    Qwen2.5 is the latest series of Qwen large language models. For Qwen2.5, we release a number of base language models and instruction-tuned language models ranging from 0.5 to 72 billion parameters. Qwen2.5 brings the following improvements upon Qwen2:

    7.6B32K ctx Apache 2.0
    6.5MSept 2024ollama
    • Same category
    • Similar size (7.6B vs 4.1B)
    • Commercial use OK
    • Widely used
    vs
  8. WeiboAI logo

    VibeThinker-3B

    WeiboAI·Chat

    ≈2.6 GB

    2.For harder math reasoning, try AMOBench , a problem set harder than the International Mathematical Olympiad (IMO), with included standard answers. Use it to evaluate VibeThinker against other SOTA models. Note: due to extreme difficulty, set max tokens to 60K–100K.

    3.1B128K ctx MIT
    2.6KJun 2026gguf
    • Same category
    • Similar size (3.1B vs 4.1B)
    • Commercial use OK
    vs
  9. CohereLabs logo

    tiny-aya-base

    CohereLabs·Chat

    ≈3.1 GB

    Cohere Labs Tiny Aya is an open weights research release of a pretrained 3.35 billion parameter model optimized for efficient, strong, and balanced multilingual representation across 70+ languages, including many lower-resourced ones.

    3.3BCC BY-NC 4.0
    2.9KFeb 2026gguf
    • Same category
    • Similar size (3.3B vs 4.1B)
    vs
  10. NVIDIA-Nemotron-3.5-Bolt-E4B-BF16 is a small language model (SLM) further post-trained from Gemma-4-E4B-it by NVIDIA. This is the full-precision (BF16) release.

    8B64K ctx openmdw-1.1
    29.7KOct 2026gguf
    • Same category
    • Similar size (8B vs 4.1B)
    • Commercial use OK
    vs
  11. allenai logo

    Olmo-3-7B-Instruct

    allenai·Chat

    ≈8.8 GB

    We introduce Olmo 3, a new family of 7B and 32B models both Instruct and Think variants. Long chain-of-thought thinking improves reasoning tasks like math and coding.

    7.3B64K ctx Apache 2.0
    128.7KNov 2025ollama
    • Same category
    • Similar size (7.3B vs 4.1B)
    • Commercial use OK
    vs
  12. openbmb logo

    BitCPM-CANN-3B

    openbmb·Chat

    ≈2.5 GB

    BitCPM-CANN is the first end-to-end 1.58-bit (ternary) large language model training system natively built on Huawei Ascend NPU.

    3B32K ctx Apache 2.0
    3KMay 2026gguf
    • Same category
    • Similar size (3B vs 4.1B)
    • Commercial use OK
    vs
  13. openbmb logo≈2.5 GB

    This is the RL initialization checkpoint used in the blog JustRL II: Scaling Small LLMs to 128K Reasoning with a Critic (中文版).

    2.8B64K ctx Unknown licence
    2.2KSept 2026
    • Same category
    • Similar size (2.8B vs 4.1B)
    vs
  14. openbmb logo

    MiniCPM5-2B-SFT

    openbmb·Chat

    ≈2.3 GB

    We are releasing MiniCPM5-2B, the second model in the MiniCPM5 series, following MiniCPM5-1B. It is a dense 2B Transformer that scales up the same training recipe, built for on-device, local deployment, and resource-constrained scenarios, reaching 2B-class open-source SOTA.

    2.5B128K ctx Apache 2.0
    6.6KAug 2026
    • Same category
    • Similar size (2.5B vs 4.1B)
    • Commercial use OK
    vs
  15. ai21labs logo

    AI21-Jamba2-3B

    ai21labs·Chat

    ≈2.4 GB

    Jamba2 3B is an ultra-compact open source model designed to bring enterprise-grade reliability to on-device deployments. At just 3B parameters, it runs efficiently on consumer devices—iPhones, Androids, Macs, and PCs—while maintaining the grounding and instruction-following capabilities required…

    3B256K ctx Apache 2.0
    6.6KJan 2026gguf
    • Same category
    • Similar size (3B vs 4.1B)
    • Commercial use OK
    vs
  16. nvidia logo≈4.2 GB

    Llama-3.1-Nemotron-Nano-4B-v1.1 is a large language model (LLM) which is a derivative of nvidia/Llama-3.1-Minitron-4B-Width-Base, which is created from Llama 3.1 8B using our LLM compression technique and offers improvements in model accuracy and efficiency.

    4.5B128K ctx NVIDIA Open Model License
    6.6KMay 2025gguf
    • Same category
    • Similar size (4.5B vs 4.1B)
    • Commercial use OK
    vs
  17. Qwen logo

    Qwen2.5-3B

    Qwen·Chat

    ≈2.6 GB

    Qwen2.5 is the latest series of Qwen large language models. For Qwen2.5, we release a number of base language models and instruction-tuned language models ranging from 0.5 to 72 billion parameters. Qwen2.5 brings the following improvements upon Qwen2:

    3.1B32K ctx Qwen Research License
    218.8KSept 2024gguf
    • Same category
    • Similar size (3.1B vs 4.1B)
    vs
  18. LiquidAI logo

    LFM3-1.7B

    LiquidAI·Chat

    ≈1.9 GB

    LFM3-1.7B is a text model for local agents and specialized applications, based on our new tri-hybrid architecture.

    1.7B256K ctx lfm1.0
    103.7KOct 2026ollama
    • Same category
    • Smaller (1.7B vs 4.1B)
    • Newer release
    vs
  19. openbmb logo≈2.3 GB

    We are releasing MiniCPM5-2B, the second model in the MiniCPM5 series, following MiniCPM5-1B. It is a dense 2B Transformer that scales up the same training recipe, built for on-device, local deployment, and resource-constrained scenarios, reaching 2B-class open-source SOTA.

    2.5B128K ctx Apache 2.0
    4.8KSept 2026
    • Same category
    • Similar size (2.5B vs 4.1B)
    • Commercial use OK
    vs
  20. CohereLabs logo

    tiny-aya-fire

    CohereLabs·Chat

    ≈3.1 GB

    Cohere Labs Tiny Aya is an open weights research release of a pretrained 3.35 billion parameter model optimized for efficient, strong, and balanced multilingual representation across 70+ languages, including many lower-resourced ones.

    3.3BCC BY-NC 4.0
    1.4KFeb 2026gguf
    • Same category
    • Similar size (3.3B vs 4.1B)
    vs
  21. CohereLabs logo

    tiny-aya-l2-thinker

    CohereLabs·Chat

    ≈3.1 GB

    Cohere Labs Tiny Aya L2-Thinker is an open-weights research release of a 3.35 billion parameter multilingual reasoning model optimized to think in the same language as the user prompt before writing the final answer.

    3.4BCC BY-NC 4.0
    330Sept 2026
    • Same category
    • Similar size (3.4B vs 4.1B)
    vs
  22. ibm-granite logo

    granite-4.2-8b

    ibm-granite·Chat

    ≈7.0 GB

    Granite is a family of open-source large language models developed by IBM, designed for enterprise and research use. Granite models are built to be versatile, safe, and efficient — covering a range of sizes and capabilities from compact edge-deployable models to large-scale reasoning systems.

    8.8B128K ctx Apache 2.0
    59.9KAug 2026ollama
    • Same category
    • Larger (8.8B vs 4.1B)
    • Commercial use OK
    vs
  23. ai9stars logo

    G9v3-3B

    ai9stars·Chat

    ≈2.7 GB

    G9v3-3B is a dense 3B causal language model from the AI9Stars team, built for local deployment and resource-constrained scenarios. It targets everyday assistant use, coding, tool-use workflows, and reasoning tasks where a compact model is preferred.

    3B128K ctx Apache 2.0
    1.2KJul 2026gguf
    • Same category
    • Similar size (3B vs 4.1B)
    • Commercial use OK
    vs
  24. allenai logo

    Olmo-3-1025-7B

    allenai·Chat

    ≈8.8 GB

    We introduce Olmo 3, a new family of 7B and 32B models. This suite includes Base, Instruct, and Think variants. The Base models were trained using a staged training approach.

    7.3B64K ctx Apache 2.0
    106.3KSept 2025gguf
    • Same category
    • Similar size (7.3B vs 4.1B)
    • Commercial use OK
    vs