{"slug": "aws-introduces-strands-decider-2b-and-strands-box-for-ai-agent-development", "title": "AWS Introduces Strands Decider 2B and Strands Box for AI Agent Development", "summary": "AWS released two open-source projects for AI agent development: Strands Decider 2B, a two-billion-parameter decision model built on Qwen3.5-2B, and Strands Box, a sandbox for controlling agent execution now in developer preview. AWS reported Strands Decider 2B at a median decision latency of roughly 115 milliseconds on an NVIDIA RTX 3090 and 153 milliseconds for small tasks on an Apple M3 MacBook, with 72.3% accuracy across 231 public tasks on the JevBench benchmark for its v19 checkpoint. Strands Box combines OS-level isolation with the open-source Dogwood policy language to enforce agent execution boundaries outside the language model.", "body_md": "AWS has introduced two open-source projects for AI agent development: [Strands Decider 2B](https://strandsagents.com/blog/introducing-strands-decider/), a small model designed to make decisions within agent workflows, and [Strands Box](https://strandsagents.com/blog/strands-box-the-big-picture/), a sandbox for controlling agent execution. Both projects address different aspects of building agents that use tools, execute code, and interact with external systems.\n\n[Strands Decider 2B](https://strandsagents.com/blog/strands-box-the-big-picture/) is a two-billion-parameter model designed to select between predefined options rather than generate lengthy text responses. It supports tasks such as model routing, tool selection, classification, guardrails, and context management. Instead of asking a general-purpose language model to generate an answer, developers can provide a set of choices and receive a decision with confidence scores.\n\nThe model is based on [Qwen3.5-2B](https://huggingface.co/Qwen/Qwen3.5-2B), with its language-model head replaced by a pointer head that scores candidate answers. It uses a [rank-16 LoRA](https://huggingface.co/blog/Neural-Hacker/lora) adapter for fine-tuning. AWS has released the model weights, training data, and scripts, allowing developers to run and modify the model locally.\n\nAccording to AWS, [Strands Decider 2B](https://strandsagents.com/blog/introducing-strands-decider/) achieved a median decision latency of approximately 115 milliseconds on an NVIDIA RTX 3090 and 153 milliseconds for small tasks on an Apple M3 MacBook. The company also evaluated the model using [JevBench](https://huggingface.co/JevBench), reporting 72.3% accuracy across 231 public tasks for its v19 checkpoint. Developers can install the strands-decider package and use its command-line interface or integrate the model into Strands agents.\n\nThe second project, [Strands Box](https://strandsagents.com/blog/strands-box-the-big-picture/), is available in developer preview and provides an isolated environment for executing AI agents. It combines operating-system-level isolation with [Dogwood](https://aws.amazon.com/blogs/opensource/introducing-dogwood-runtime-verification-for-ai-agents/), an open-source policy language that defines which actions an agent can perform. The sandbox supports shell commands, Python execution, MCP tools, filesystem operations, and network access.\n\n[Strands Box](https://strandsagents.com/blog/strands-box-the-big-picture/) uses kernel-enforced isolation to route agent interactions through a policy layer. Developers can allow or deny actions based on their content and execution history. For example, a policy can require an agent to pass tests before pushing code to a repository or restrict the frequency of external API calls. These controls operate outside the language model, reducing reliance on prompt instructions to enforce execution boundaries.\n\nAWS says [Strands Box](https://strandsagents.com/blog/strands-box-the-big-picture/) can work with different models and agent frameworks, whether running locally or in cloud environments. This allows developers to introduce execution policies without replacing their existing agent architecture.\n\nStrands Decider 2B competes with [TypeSafe AI's Jev](https://www.infoq.com/news/2026/10/typesafe-ai-jev-released/), another specialized model designed to make structured decisions without generating lengthy responses. Both support classification, routing, and tool selection, returning confidence scores alongside their decisions. The main difference is deployment: Strands Decider 2B is open source and can run locally, while Jev is available through a managed API. AWS evaluated Strands Decider using JevBench, although the published results do not establish a direct performance advantage over Jev.\n\nStrands Box addresses problems also handled by container-based sandboxes, [Firecracker microVMs](https://firecracker-microvm.github.io/), and hosted agent execution environments. While conventional sandboxing primarily restricts filesystem, process, and network access, Strands Box adds policies based on tool arguments and previous actions. This allows developers to define conditions that must be satisfied before an agent executes specific operations.\n\nEarly developer feedback has focused on the practical limitations of integrating small decision models into agent workflows. In a hands-on tutorial, [AWS Community Builder Yuuki Yamashita](https://builder.aws.com/content/3K6mQXLgor1NkkxPjNR5A6GlFFg/strands-decider-2b-a-hands-on-a-13-year-old-can-follow) reported that:\n\nStrands Decider 2B could prevent an agent from making an unsupported weather-tool call.\n\n[GitHub discussions](https://github.com/strands-labs/strands-decider/issues) have also identified issues involving concurrent requests, Apple Silicon compatibility, and cold-start latency, highlighting areas developers may need to evaluate before using the model in production.\n\nAI Builder [Dipen Vaja](https://x.com/dipenvaja/status/2109788021205917816) commented on X:\n\nI'd rather an agent hit a wall mid-run than finish a bad path. AWS open sourced Strands Box today: a kernel sandbox that routes shell, Python, tools and network through one policy layer, so a rule can block outbound HTTP after the agent reads a customer file.\n\nTogether, Strands Decider 2B and Strands Box extend the [Strands ecosystem](https://www.infoq.com/news/2026/03/aws-strands-agents/) beyond agent orchestration into decision-making and execution control. The projects provide separate components for selecting agent actions and enforcing restrictions on how those actions are executed.", "url": "https://wpnews.pro/news/aws-introduces-strands-decider-2b-and-strands-box-for-ai-agent-development", "canonical_source": "https://www.infoq.com/news/2026/10/aws-strands-agents/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=global", "published_at": "2026-10-16 10:47:00+00:00", "updated_at": "2026-10-16 11:12:57.884028+00:00", "lang": "en", "topics": ["ai-agents", "ai-tools", "large-language-models", "ai-safety", "developer-tools"], "entities": ["AWS", "Strands Decider 2B", "Strands Box", "Qwen3.5-2B", "Dogwood", "JevBench", "TypeSafe AI", "Jev"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/aws-introduces-strands-decider-2b-and-strands-box-for-ai-agent-development", "markdown": "https://wpnews.pro/news/aws-introduces-strands-decider-2b-and-strands-box-for-ai-agent-development.md", "text": "https://wpnews.pro/news/aws-introduces-strands-decider-2b-and-strands-box-for-ai-agent-development.txt", "jsonld": "https://wpnews.pro/news/aws-introduces-strands-decider-2b-and-strands-box-for-ai-agent-development.jsonld"}}