@GitTrend0x: A pure local desktop automation powerhouse, and most importantly, saves money! https://github.com/microsoft/fara This is Fara-7B, an efficient Computer Use Agent small model from Microsoft! In a word, it surpasses traditional large model CUA: only 7B parameters...

A dedicated tool for pure local desktop automation, and most importantly, it saves money! https://github.com/microsoft/fara This is Fara-7B, Microsoft’s efficient Computer Use Agent small model! In a word, it outperforms traditional large CUA models: with only 7B parameters, it surpasses larger models on multiple real-world web task benchmarks. It directly uses visual perception to operate on web page coordinates (click, type, scroll), averaging only ~16 steps to complete tasks. It perfectly supports local deployment (Ollama/vLLM/LM Studio), giving your agent truly low-cost, high-efficiency computer use capabilities!

— # microsoft/fara Source: https://github.com/microsoft/fara # Fara-7B: An Efficient Agentic Model for Computer Use Microsoft (https://aka.ms/msaif/fara) Hugging Face Model (https://huggingface.co/microsoft/Fara-7b) Foundry (https://aka.ms/foundry-fara-7b) Dataset (https://huggingface.co/datasets/microsoft/WebTailBench) Dataset (https://huggingface.co/datasets/microsoft/CUAVerifierBench) Paper (https://arxiv.org/abs/2511.19663) — ## Updates * 2026-05-21 - Fara1.5 agent harness coming soon! * 2026-05-12 — Refreshed WebTailBench (V2) tasks and rubrics. Many V1 tasks had calendar-bound dates that expired (Nov 2025); V2 rolls those forward and revises the precomputed rubrics for the full 609-task suite. Available now as the test_v2 split on microsoft/WebTailBench (https://huggingface.co/datasets/microsoft/WebTailBench). A side-by-side V1↔V2 diff (task strings and rubric JSON) is hosted here (https://microsoft.github.io/fara/docs/webtailbench_v1_v2_diff.html). * 2026-04-19 — Released CUAVerifierBench (https://huggingface.co/datasets/microsoft/CUAVerifierBench), a human-annotated benchmark for evaluating CUA verifiers (i.e. judges that score agent trajectories). Two splits — fara7b_om2w_browserbase (106 Fara-7B Online-Mind2Web/Browserbase trajectories, ~2 reviewers each) and internal (154 trajectories from a heldout aurora-v2 task suite) — with per-judge UV-blind / UV-informed labels, Universal Verifier outputs, and legacy verifier outputs side-by-side. The build script that produced the dataset lives alongside the data on HuggingFace. * 2026-04-18 — Removed the autogen-core / autogen-ext dependency from webeval; chat completion clients are now self-contained under webeval/src/webeval/oai_clients/. No more autogen submodule install step; just pip install -e .[vllm] then cd webeval; pip install -e .. * 2026-04-18 — Incorporated WebTailBench (initial / now-stale version) directly into the repo as a first-class benchmark. The loader auto-downloads WebTailBench-v1-rubrics.tsv from microsoft/WebTailBench (https://huggingface.co/datasets/microsoft/WebTailBench) and threads each task’s published precomputed_rubric through to the verifier. Reproducibility CLI lives in webeval/scripts/webtailbench.py. * 2026-04-18 — Released the Universal Verifier (MMRubricAgent) as the official verifier for WebTailBench. Multimodal, rubric-grounded, two-model ensemble (gpt-5.2 + o4-mini) with per-criterion scoring, outcome verification, and first-point-of-failure analysis. A stand-alone parallel runner is at webeval/scripts/verify_trajectories.py for re-scoring any directory of webeval trajectories without touching the solver. — ## Overview Fara-7B is Microsoft’s first agentic small language model (SLM) designed specifically for computer use. With only 7 billion parameters, Fara-7B is an ultra-compact Computer Use Agent (CUA) that achieves state-of-the-art performance within its size class and is competitive with larger, more resource-intensive agentic systems. Try Fara-7B locally as follows (see Installation for detailed instructions on Windows ) or via Magentic-UI: bash # 1. Clone repository git clone https://github.com/microsoft/fara.git cd fara # 2. Setup environment python3 -m venv .venv source .venv/bin/activate pip install -e . playwright install Then in one process, host the model: bash vllm serve "microsoft/Fara-7B" --port 5000 --dtype auto Then you can iteratively query it with: bash fara-cli --task "whats the weather in new york now" To try Fara-7B inside Magentic-UI, please follow the instructions here Magentic-UI + Fara-7B (https://github.com/microsoft/magentic-ui/blob/main/README.md#fara-7b). You will need to serve the model as before, but instead of fara-cli you can use Magentic-UI which has a nice UI (see video demos below). Notes: - If you’re using Windows, we highly recommend using WSL2 (Windows Subsystem for Linux). Please see the Windows instructions in the Installation section. - You might need to do --tensor-parallel-size 2 with vllm command if you run out of memory Shopping GitHub Issues Directions with Cheese ### What Makes Fara-7B Unique Unlike traditional chat models that generate text-based responses, Fara-7B leverages computer interfaces—mouse and keyboard—to perform multi-step tasks on behalf of users. The model: - Operates visually by perceiving webpages and taking actions like scrolling, typing, and clicking on directly predicted coordinates without accessibility trees or separate parsing models - Enables on-device deployment due to its compact 7B parameter size, resulting in reduced latency and improved privacy as user data remains local - Completes tasks efficiently, averaging only ~16 steps per task compared to ~41 for comparable models Fara-7B is trained using a novel synthetic data generation pipeline built on the Magentic-One (https://www.microsoft.com/en-us/research/articles/magentic-one-a-generalist-multi-agent-system-for-solving-complex-tasks/) multi-agent framework, with 145K trajectories covering diverse websites, task types, and difficulty levels. The model is based on Qwen2.5-VL-7B (https://arxiv.org/abs/2502.13923) and trained with supervised fine-tuning. ### Key Capabilities Fara-7B can automate everyday web tasks including: - Searching for information and summarizing results - Filling out forms and managing accounts - Booking travel, movie tickets, and restaurant reservations - Shopping and comparing prices across retailers - Finding job postings and real estate listings ### Performance Highlights Fara-7B achieves state-of-the-art results across multiple web agent benchmarks, outperforming both comparable-sized models and larger systems: | Model | Params | WebVoyager | Online-M2W | DeepShop | WebTailBench | |—––|––––|————|————|–––––|–––––––| | SoM Agents | | | | | | | SoM Agent (GPT-4o-0513) | - | 90.6 | 57.7 | 49.1 | 60.4 | | SoM Agent (o3-mini) | - | 79.3 | 55.4 | 49.7 | 52.7 | | SoM Agent (GPT-4o) | - | 65.1 | 34.6 | 16.0 | 30.8 | | GLM-4.1V-9B-Thinking | 9B | 66.8 | 33.9 | 32.0 | 22.4 | | Computer Use Models | | | | | | | OpenAI computer-use-preview | - | 70.9 | 42.9 | 24.7 | 25.7 | | UI-TARS-1.5-7B | 7B | 66.4 | 31.3 | 11.6 | 19.5 | | Fara-7B | 7B | 73.5 | 34.1 | 26.2 | 38.4 | Table: Online agent evaluation results showing success rates (%) across four web benchmarks. Results are averaged over 3 runs. ### WebTailBench: A New Benchmark for Real-World Web Tasks We are releasing WebTailBench (https://huggingface.co/datasets/microsoft/WebTailBench), a new evaluation benchmark focusing on 11 real-world task types that are underrepresented or missing in existing benchmarks. The benchmark includes 609 tasks across diverse categories, with the first 8 segments testing single skills or objectives (usually on a single website), and the remaining 3 evaluating more difficult multi-step or cross-site tasks. #### WebTailBench Detailed Results | Task Segment | Tasks | SoM GPT-4o-0513 | SoM o3-mini | SoM GPT-4o | GLM-4.1V-9B | OAI Comp-Use | UI-TARS-1.5 | Fara-7B | |–––––––|—––|—————–|———––|————|———––|–––––––|———––|———––| | Single-Site Tasks | | Shopping | 56 | 62.5 | 71.4 | 38.1 | 31.0 | 42.3 | 41.1 | 52.4 | | Flights | 51 | 60.1 | 39.2 | 11.1 | 10.5 | 17.6 | 10.5 | 37.9 | | Hotels | 52 | 68.6 | 56.4 | 31.4 | 19.9 | 26.9 | 35.3 | 53.8 | | Restaurants | 52 | 67.9 | 59.6 | 47.4 | 32.1 | 35.9 | 22.4 | 47.4 | | Activities | 80 | 70.4 | 62.9 | 41.7 | 26.3 | 30.4 | 9.6 | 36.3 | | Ticketing | 57 | 58.5 | 56.7 | 37.4 | 35.7 | 49.7 | 30.4 | 38.6 | | Real Estate | 48 | 34.0 | 17.4 | 20.1 | 16.0 | 9.0 | 9.7 | 23.6 | | Jobs/Careers | 50 | 49.3 | 44.0 | 32.7 | 22.7 | 20.7 | 20.7 | 28.0 | | Multi-Step Tasks | | Shopping List (2 items) | 51 | 66.0 | 62.7 | 17.0 | 7.8 | 34.0 | 20.9 | 49.0 | | Comparison Shopping | 57 | 67.3 | 59.1 | 27.5 | 22.8 | 1.2 | 8.8 | 32.7 | | Compositional Tasks | 55 | 51.5 | 39.4 | 26.7 | 17.0 | 10.3 | 9.1 | 23.0 | | Overall | | Macro Average | 609 | 59.7 | 51.7 | 30.1 | 22.0 | 25.3 | 19.9 | 38.4 | | Micro Average | 609 | 60.4 | 52.7 | 30.8 | 22.4 | 25.7 | 19.5 | 38.4 | Table: Breakdown of WebTailBench results across all 11 segments. Success rates (%) are averaged over 3 independent runs. Fara-7B achieves the highest performance among computer-use models across all task categories. Coming Soon: - Task Verification pipeline for LLM-as-a-judge evaluation - Official human annotations of WebTailBench (in partnership with BrowserBase) ### CUAVerifierBench: Evaluating the Verifiers Themselves While WebTailBench measures agents, CUAVerifierBench (https://huggingface.co/datasets/microsoft/CUAVerifierBench) measures the judges that score those agents. Each row pairs a Fara-7B agent trajectory (instruction, screenshots, web_surfer log, final answer) with one human reviewer’s verdict, plus the verdicts produced by the Universal Verifier (MMRubricAgent) and several legacy verifiers — so researchers can compute verifier–human agreement (Cohen’s κ, accuracy, F1) on a fixed corpus and iterate on new judge prompts / architectures against a frozen ground-truth set. The dataset is exposed as two HuggingFace configs joinable on task_id: | Config | Granularity | Contents | |—|—|—| | trajectories | one row per task | instruction, screenshots, web_surfer log, verifier outputs, task-level human aggregates | | annotations | one row per (task, judge) | per-reviewer outcome / process labels and free-text justifications | Two splits ship today: | Split | Source | Trajectories | Annotation rows | |—|—|—|—| | fara7b_om2w_browserbase | Fara-7B trajectories on Online-Mind2Web tasks executed via Browserbase | 106 | 215 (≈2 reviewers/task; UV-blind and UV-informed stages) | | internal | Heldout aurora-v2 task suite scored with the same WebSurfer + verifier stack | 154 | 154 (1 reviewer/task; UV-blind only) | Reviewer identities are anonymized as Judge1 … JudgeN using a single shared map across both splits. The build script that produced the dataset (with full schema + provenance) ships alongside the data on HuggingFace at microsoft/CUAVerifierBench (https://huggingface.co/datasets/microsoft/CUAVerifierBench); see the dataset README (https://huggingface.co/datasets/microsoft/CUAVerifierBench/blob/main/README.md) for the full column list. python from datasets import load_dataset trajs = load_dataset("microsoft/CUAVerifierBench", "trajectories", split="fara7b_om2w_browserbase") anns = load_dataset("microsoft/CUAVerifierBench", "annotations", split="fara7b_om2w_browserbase") ### Evaluation Infrastructure Our evaluation setup leverages: 1. Playwright - A cross-browser automation framework that replicates browser environments 2. Abstract Web Agent Interface - Allows integration of any model from any source into the evaluation environment 3. Fara-Agent Class - Reference implementation for running the Fara model > Note: Fara-7B is an experimental release designed to invite hands-on exploration and feedback from the community. We recommend running it in a sandboxed environment, monitoring its execution, and avoiding sensitive data or high-risk domains. — # Installation ## Linux The following instructions are for Linux systems, see the Windows section below for Windows instructions. Install the package using pip and set up the environment with Playwright: bash # 1. Clone repository git clone https://github.com/microsoft/fara.git cd fara # 2. Setup environment python3 -m venv .venv source .venv/bin/activate pip install -e .[vllm] playwright install Note: If you plan on hosting with Azure Foundry only, you can skip the [vllm] and just do pip install -e . ## Windows For Windows, we highly recommend using WSL2 (Windows Subsystem for Linux) to provide a Linux-like environment. However, if you prefer to run natively on Windows, follow these steps: bash # 1. Clone repository git clone https://github.com/microsoft/fara.git cd fara # 2. Setup environment python3 -m venv .venv .venv\Scripts\activate pip install -e . python3 -m playwright install ## Hosting the Model Recommended: The easiest way to get started is using Azure Foundry hosting, which requires no GPU hardware or model downloads. Alternatively, you can self-host with vLLM if you have GPU resources available. ### Azure Foundry Hosting (Recommended) Deploy Fara-7B on Azure Foundry (https://ai.azure.com/explore/models/Fara-7B/version/2/registry/azureml-msr) without needing to download weights or manage GPU infrastructure. Setup: 1. Deploy the Fara-7B model on Azure Foundry and obtain your endpoint URL and API key Then create a endpoint configuration JSON file (e.g., azure_foundry_config.json): json { "model": "Fara-7B", "base_url": "https://your-endpoint.inference.ml.azure.com/", "api_key": "YOUR_API_KEY_HERE" } Then you can run Fara-7B using this endpoint configuration. 2. Run the Fara agent: bash fara-cli --task "how many pages does wikipedia have" --endpoint_config azure_foundry_config.json [--headful] Note: you can also specify the endpoint config with the args --base_url [your_base_url] --api_key [your_api_key] --model [your_model_name] instead of using a config JSON file. Note: If you see an error that the fara-cli command is not found, then try: bash python -m fara.run_fara --task "what is the weather in new york now" That’s it! No GPU or model downloads required. ### Self-hosting with vLLM or LM Studio / Ollama If you have access to GPU resources, you can self-host Fara-7B using vLLM. This requires a GPU machine with sufficient VRAM (e.g., 24GB or more). Only on Linux: all that is required is to run the following command to start the vLLM server: bash vllm serve "microsoft/Fara-7B" --port 5000 --dtype auto For quantized models or lower VRAM GPUs, please see Fara-7B GGUF on HuggingFace (https://huggingface.co/bartowski/microsoft_Fara-7B-GGUF). For Windows/Mac, vLLM is not natively supported. You can use WSL2 on Windows to run the above command or LM Studio / Ollama as described below. Otherwise, you can use LM Studio (https://lmstudio.ai/) or Ollama (https://ollama.com/) to host the model locally. We currently recommend the following GGUF versions of our models Fara-7B GGUF on HuggingFace (https://huggingface.co/bartowski/microsoft_Fara-7B-GGUF) for use with LM Studio or Ollama. Select the largest model that fits your GPU. Please ensure that context length is set to at least 15000 tokens and temperature to 0 for best results. Then you can run Fara-7B pointing to your local server: Run the test script to see Fara in action: bash fara-cli --task "what is the weather in new york now" If you didn’t use vLLM to host, please specify the correct --base_url [your_base_url] --api_key [your_api_key] --model [your_model_name] If you see an error that the fara-cli command is not found, then try: bash python -m fara.run_fara --task "what is the weather in new york now" # Reproducibility We provide a framework in webeval/ to reproduce our results on WebVoyager and OnlineMind2Web. Agentic evaluations on live websites present unique challenges due to day-to-day changes. We implement several measures to ensure reliable and comparable evaluations: BrowserBase Integration We employ BrowserBase to manage browser session hosting, enabling reliable browser instance management. Time-sensitive Task Updates Tasks in benchmarks like WebVoy