DEV Community

SparkLLM
SparkLLM

Posted on

Spark X2.5-4B & 1.7B: the only on-device models with native 1M-token context — now open source

#opensource #llm #edgeai #ondevice

Today SparkLLM releases and open-sources two on-device general models: Spark X2.5-4B and Spark X2.5-1.7B. Both natively support a context window of up to 1,000,000 tokens — as far as we know, the only on-device models to do so.

Why 1M context on-device

In real work, you rarely hand a model a single question — you hand it a whole after-sales manual, a set of meeting materials, a batch of project docs, or an entire code repository. On-device models used to chop long content into pieces and ask about each separately, which loses context and drops information.

Spark X2.5-4B and 1.7B natively support up to a 1M-token context window, trained on hundreds-of-billions-of-tokens of high-quality long-document data, so they can take in and reason over far more information in a single task — and keep the full picture across a continuous, multi-step interaction.

Not just answering — doing the work

Long context decides whether the model can see everything; agent + tool-use decides whether it can act on it.

  • Office (with Loomy): upload a sales spreadsheet and ask for an analysis plus a bilingual department report — X2.5-4B writes a script to aggregate the data, extracts key metrics and trends, generates a ~3,000-word Chinese report, produces an English version in the same structure, and validates content, structure and layout end to end.
  • Code: on algorithm implementation, completion and generation, X2.5-4B rivals cloud models 2–3× its size. It plugs into open harnesses like DeepSeek Harness, OpenCode, Codex and Pi for local dev and automation — with low latency, offline use, and code kept on-device.
  • Smart home: on the Domux smart-home test set, X2.5-1.7B reaches 90.3% end-to-end command accuracy at 0.85s average latency.
  • Robotics: both sizes suit continuous perception-and-execution on-robot or on edge devices — operation control, target tracking, navigation decisions — with less dependence on the cloud.

Domestic compute, open deployment

Both models were trained end to end on a fully domestic compute platform, pre-trained on ~20T tokens of diverse data and refined with high-quality SFT and RL. They run on NVIDIA, Huawei, Hygon and Houmo hardware, are compatible with vLLM, SGLang and llama.cpp, deploy quickly via Ollama and LM Studio, and support incremental training with LLaMA-Factory.

Get it

Weights, code and deployment docs are open now on GitHub and Hugging Face, and the API is live on iFlytek Xingchen MaaS (free for a limited time):

And on Sep 7, we'll release Spark X2.5-293B, further upgrading code and agent capabilities. Stay tuned.

Figures above are from the official release.

Top comments (0)