SparkLLM has released Spark-X2.6-4B, a new compact language model designed for on-device agents and real-world applications. This model builds upon its predecessor, Spark-X2.5-4B, with enhanced post-training capabilities that improve performance in coding, agent workflows, long-context tasks, multilingual use, and reasoning. Spark-X2.6-4B offers controllable thinking levels to balance reasoning effort, latency, and cost, and supports on-device inference across various hardware and runtimes like RTX 4090 and Mac M5 Pro. AI
IMPACT Enhances on-device AI capabilities, particularly for agents and coding tasks, potentially lowering barriers for local AI deployment.
RANK_REASON Frontier-lab model release with system card. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
- Apache Software License 2.0
- llama.cpp
- Mac M5 Pro
- Mlx
- Ollama
- RTX 4090
- SGLang
- SparkLLM
- SparkLLM/Spark-X2.6-4B
- Spark-X2.6-4B
- tensorrt
- transformers
- vLLM
- XHToken/Spark-X2.6-4B
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →