Running AWS Strands Decider 2B Locally: A Complete Setup Guide for AI Routing & Multi-RAG Systems

Description and use

Deploying AWS Strands Decider 2B locally on Windows via WSL2 separates routing decisions from LLM generation in GenAI workflows, improving decision reliability.

Problem

GenAI applications need reliable pre‑LLM decisions for routing, scoring, and orchestration, but traditional LLMs handle both decision making and response generation.

Outcome

Local deployment of Strands Decider 2B enables efficient routing decisions in GenAI pipelines, separating decision logic from LLM generation.

Workflow

• Verify WSL2 installation with wsl -l -v
• Create workspace at /mnt/c/GENAI/strands
• Update Ubuntu and install dependencies (sudo apt update && sudo apt install -y python3 python3-pip python3-venv python3-dev build-essential gcc g++)
• Create and activate Python virtual environment (python3 -m venv .venv && source .venv/bin/activate)
• Upgrade pip and install packages (pip install --upgrade pip setuptools wheel)
• Install Strands Decider (pip install strands-decider)
• Verify installation (strands-decider --help)
• Install Hugging Face Hub (pip install huggingface_hub)
• List available models (python -c "from huggingface_hub import list_models; [print(m.id) for m in list_models(search='strands')]")
• Start model server (strands-decider serve StrandsAgents/strands-decider-2B-hobson-v19 --device cpu)
• Verify API via browser or curl (http://127.0.0.1:8000/docs or curl http://127.0.0.1:8000/openapi.json)
• Create Python demo script to send decision requests

Evidence

AWS introduced Strands Decider 2B, a lightweight decision model designed specifically for routing, classification, scoring, and orchestrating agent workflows.

Source

strands-decider serve StrandsAgents/strands-decider-2B-hobson-v19 --device cpu

Source

Application startup complete. Uvicorn running on http://127.0.0.1:8000

Source