Oriel — AI & technology worth noticing, every day.

XHToken/Spark-X2.6-4B-GGUF

AI model · https://huggingface.co/XHToken/Spark-X2.6-4B-GGUF

First seen by Oriel on .

AI summary

Spark-X2.6-4B is an Apache 2.0-licensed, 4B-parameter on-device model whose post-training improves coding, agentic tasks, long-context handling, multilingual ability, and reasoning without changing its architecture.

  • The model offers none, low, and high thinking levels so applications can trade off reasoning quality, latency, and inference cost.
  • The repository reports best-in-class results among similarly sized models and performance matching or exceeding some 9B models, but the provided text includes no benchmark scores.
  • It supports Transformers, SGLang, vLLM, MLX, llama.cpp, Ollama, LM Studio, and Unsloth, with deployment guidance for NVIDIA GPUs and Ascend NPUs.
  • SGLang examples configure a context length of 1,048,576 tokens, which requires sufficient device memory and can be reduced when necessary.
  • Fine-tuning is recommended through Llama-Factory; llama.cpp native spark2_5 support requires version b10828 or later.

#language models #on-device ai #llm agents #gguf

Generated by gpt-5.6-sol. Check the original for details.

Metric trends

Last 30 days · Up to 120 recorded samples per source and metric, with reposts kept separate · First and latest retained

Likes

Total likes reported by the source.

  • Hugging Face Models: 263, latest observation ; +158 over observed period since

Downloads (30d)

Rolling 30-day downloads. A change in this value is not new downloads during the observed period.

  • Hugging Face Models: 317,709, latest observation ; +307,885 over observed period since

Aliases

SourceExternal IDURL
Hugging Face ModelsXHToken/Spark-X2.6-4B-GGUFhttps://huggingface.co/XHToken/Spark-X2.6-4B-GGUF Read discussion