AI BriefingAgents-A1-4B Model Released
Key point
InternScience has released Agents-A1, a 4B-scale model that maximizes agentic task performance.
Details
InternScience announced Agents-A1-4B, a model that maximizes reasoning ability as an agent instead of scaling up parameter size.
This model is based on Qwen3.5-4B, and despite being a small-scale model, it has proven strong performance across various benchmarks.
Key performance metrics:
- GAIA (General AI Assistants): Scored 96.0, showing overwhelming agentic task performance compared to models of the same class
- IFEval (Instruction Following): Scored 94.8, demonstrating high instruction-following capability
- XBench-DS-2510: Scored 86.0, proving search and reasoning performance
- SciCode: Scored 44.3, demonstrating efficiency in engineering and research tasks
The model's weights, including a GGUF format, have been released via Hugging Face, making it usable across a variety of environments.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Apple Releases LensVLM-9B for Processing Compressed Text Images
Qwen Releases 'RecreationWorld' for Training AI Agents to Rebuild AppsDeepSeek Releases DSec: Elastic Sandbox Infrastructure for Large-Scale Agentic Training