Microsoft Research has released Orchard, an open-source framework for building, training, and evaluating autonomous AI agents that the company says lets relatively small open-weight models approach the performance of frontier systems more than ten times their size. The project, detailed in a blog post on August 3, 2026, is the company's most ambitious bid yet to give the broader research community the kind of infrastructure that has, until now, largely remained locked inside proprietary labs.
The release arrives as the AI industry shifts from static question-answering toward agents that can plan, reason, and act across multistep environments. For more on how the field is moving toward autonomous systems, see our latest AI industry coverage.
What Orchard actually is
At the center of the project is Orchard Env, a lightweight, Kubernetes-based environment service that provides reusable, isolated components for running agents at scale. According to Microsoft, the same service can support software-engineering agents, web-browsing agents, and personal-assistant agents without modification. It handles everything from collecting training data to reinforcement-learning rollouts and evaluation.
The key architectural decision is that the runtime environment is treated as a standalone, reusable service rather than infrastructure embedded inside a specific training framework. Because Orchard Env sits on Kubernetes, it can create, manage, and remove thousands of isolated containers in parallel. Teams can introduce new benchmarks, agent systems, or training algorithms without rebuilding the underlying infrastructure from scratch.
Training inside real harnesses
One of Orchard's distinguishing features is its ability to train agents directly inside the deployment harnesses they will actually use in production. Today's most capable agents rarely run as a bare model. They operate through sophisticated harnesses such as Codex, OpenClaw, and ZeroClaw that manage multi-turn reasoning, tool use, and connections to external systems.
Open training tools usually cannot handle these stateful, multi-process harnesses, forcing researchers to train on a simplified stand-in and then deploy in the real setting, which creates a mismatch between training and deployment. Orchard closes this gap: a lightweight proxy records the harness's own model calls as training data while each rollout runs in its own container, allowing an agent to be trained end-to-end directly in the harness it will use in production.
Three domain-specific recipes
To demonstrate the approach, Microsoft released three training workflows.
Orchard-SWE: software engineering
Orchard-SWE targets one of the most demanding settings for autonomous agents: multi-step reasoning over real codebases. It was built using the Mini-SWE-Agent framework and evaluated on SWE-bench Verified, a benchmark that tests a model's ability to navigate, diagnose, and repair real-world codebases.
To train the system, Microsoft distilled 107,000 agent interactions from two advanced open-weight models, MiniMax-M2.5 and Qwen3.5-397B, covering a broad range of GitHub issues. Using a technique called credit-assignment supervised fine-tuning, the system learns from the productive portions of partial attempts rather than discarding them entirely.
The results move the model from a 61.4% baseline on SWE-bench Verified to 69.1% with reinforcement learning, and 69.7% with dense-reward techniques. With value-model reranking, the figure reaches 73.0% — a new state of the art among open-source models of comparable size, using only about 3 billion active parameters.
Orchard-GUI: web navigation
Orchard-GUI trains a 4-billion-parameter vision-language model as a browser agent. Despite using a relatively small amount of supervision — 400 distilled demonstrations combined with 2,200 open-ended training tasks — the model achieved 74.1% on WebVoyager, 67.0% on Online-Mind2Web, and 64.0% on DeepShop, for an average of 68.4%. Microsoft says these results place it among the strongest open-source web agents to date.
Orchard-Claw: personal assistant tasks
Orchard-Claw фокусируется на повседневных задачах продуктивности, таких как чтение и составление электронных писем, управление календарями и поиск информации. Обученный всего на 200 синтетических задачах и оцененный по тесту Claw-Eval, он выполнил 59,6% задач при трех попытках, а в сочетании с более мощной системой агентов ZeroClaw этот показатель увеличился до 73,9%.
Почему важны результаты малых моделей
Результаты тестов примечательны тем, что они получены от моделей примерно с 3–4 миллиардами активных параметров — гораздо меньше, чем передовые системы OpenAI, Anthropic и Google, которые доминируют в таблицах лидеров. Аргумент Microsoft заключается в том, что правильная инфраструктура и среда обучения могут устранить большую часть разрыва, сделав работоспособные агентные системы доступными для исследователей и организаций, которые не могут позволить себе обучение или запуск крупнейших запатентованных моделей.
Наряду с моделями и рабочими процессами Microsoft опубликовала данные обучения и методы оценки, использованные для их построения, с заявленной целью помочь более широкому исследовательскому сообществу создавать и изучать открытые агентные системы. Работу возглавляли Баолинь Пэн, Вэньлинь Яо, Цяньхуэй Ву, Хао Чэн и Цзяньфэн Гао из Microsoft Research.
Будьте впереди ИИ
Релиз Microsoft Orchard сигнализирует о том, что инфраструктура, лежащая в основе передового агентного ИИ, начинает демократизироваться. Узнайте больше последние новости AI и Читать больше новостей AI →
