SparkLLM open-sources X2.5 4B/1.7B on-device models with native 1M-token context
vllm_project · x · 2026-09-01
SparkLLM has open-sourced Spark-X2.5-4B and X2.5-1.7B, compact on-device agent models claimed to be the only ones with native context up to 1,000,000 tokens. Built on a hybrid-attention architecture and tuned for agents, code, math and instruction following, they support 200+ languages. The vLLM project confirmed day-0 support for both via Spark's out-of-tree general plugin, so they can be served with vLLM from launch day.
Related event: SparkLLM Open-Sources X2.5 On-Device Models with 1M Context(2 posts)→
More from coding & agent
- Open-source virtual computer MCP gives LLMs a full desktop in your browser — Ambitious-Prompt-975 · 2026-09-01
- Dev builds FinBridge MCP: 33 tools for Korean + US market data in agent workflows — JakeChj · 2026-09-01
- Dev Reports Sol Built a Useful Model Probe Tool and Handled Refactors Well — voooooogel · 2026-09-01
- Indie founder rebuilds startup site in 3 days with AI: Sol does grunt work, human directs — kylegawley · 2026-09-01
- mitsuhiko: server-side context compaction breaks session portability for no technical reason — mitsuhiko · 2026-09-01
- ElevenLabs launches MCP integration for Claude — petewoodbridge · 2026-09-01