yasserrmd/AgenticCoder-4B
Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:4BQuant:BF16Ctx Length:32kPublished:Jul 29, 2025Architecture:Transformer0.0K Warm

yasserrmd/AgenticCoder-4B is a 4 billion parameter language model, merging Menlo/Jan-nano and ertghiu256/qwen3-4b-code-reasoning, designed for autonomous agent workflows and intelligent code reasoning. It excels in multi-step planning, task decomposition, and robust Python code generation, explanation, and optimization. This compact model is optimized for real-world assistant scenarios, research agents, and smart development tools, supporting tool interaction simulation for tasks like CSV analysis and OCR.

Loading preview...

AgenticCoder-4B: Compact Agentic & Code Reasoning Model

AgenticCoder-4B is a 4 billion parameter language model developed by yasserrmd, specifically engineered for autonomous agent workflows and intelligent code reasoning. It is a merge of Menlo/Jan-nano (for planning and tool-use) and ertghiu256/qwen3-4b-code-reasoning (for coding and logic), resulting in a balanced model for complex tasks.

Key Capabilities

  • Agentic Planning & Multi-Context Processing (MCP) Alignment: Optimized for multi-step reasoning, task decomposition, and memory-contextual workflows, enabling sophisticated agentic behavior.
  • Code Understanding & Reasoning: Demonstrates strong capabilities in Python code generation, script explanation, optimization, and multi-turn task development.
  • Tool Use Simulation: Handles realistic prompts involving tool interaction, such as CSV analysis, OCR, and file parsing within code contexts.
  • Compact & Efficient: At 4 billion parameters, it is lightweight, making it suitable for cost-efficient deployment, edge device integration, and further fine-tuning.

Good For

  • Real-world Assistant Scenarios: Ideal for building intelligent assistants that require planning and code execution.
  • Research Agents: Supports autonomous research tasks requiring logical reasoning and data processing.
  • Smart Development Tools: Can be integrated into tools for code generation, optimization, and task automation.
  • Specific Examples: Designing Python curricula, writing non-recursive JSON scanning functions, extracting text from images via OCR, and summarizing/visualizing data from CSVs.