Create a private Ultralytics workspace project by specifying a display name and URL slug, with optional owner, visibility, or description settings to organize your assets.
MIT
650,527 tools. Updated 2026-10-14 06:12
"Ultralytics" matching MCP tools:
- Create a dataset in your Ultralytics workspace with configurable name, slug, task, classes, owner, and visibility; defaults to private unless public is set.MIT
- Search public Ultralytics Explore datasets by keyword, task, or sort order to find and filter datasets for training, detection, or segmentation projects.MIT
- Search public projects on Ultralytics Explore by keyword, with sorting and pagination to refine results.MIT
- List model deployment records in your Ultralytics workspace to track and manage active and past deployments.MIT
- Lists computer-vision projects in your Ultralytics workspace. Filter by owner and set result limits to browse your projects.MIT
Matching MCP Servers
- AlicenseBqualityAmaintenanceEnables management and interaction with the Ultralytics Platform, including browsing projects/datasets/models, starting training, running predictions, and creating exports.44434 npm10MIT
- FlicenseNot gradedqualityDmaintenanceA Model Context Protocol compliant server that provides RESTful API access to Ultralytics YOLO operations for computer vision tasks including training, validation, prediction, export, tracking, and benchmarking.4-
- List datasets in your Ultralytics workspace, optionally filtered by workspace owner or username, to browse available data for projects, training, or predictions.MIT
- Report a model's evaluation plots (PR, F1, precision, recall, confusion matrix) by listing each plot's field shapes, or fetch one plot's full data by type.MIT
- Start a billable auto-annotation run that labels a dataset with a chosen model; requires confirm_cost, snapshots a dataset version, and can be polled or undone.MIT
- Restore a dataset to a saved version by version number to undo later changes, replacing current images, labels, and splits with that snapshot.MITDestructive
- Stop an active deployment by owner/slug or bare slug to cut costs while preserving its URL and configuration. Note: the deployment still counts toward quota and cannot be restarted from this tool.MITDestructive
- Launch a cloud training job for YOLO models on one dataset, validating it upfront to prevent wasted credits. Requires cost confirmation and may overwrite training history.MITDestructive
- Stop or dismiss a dataset's auto-annotation run by slug or URI. Checks status first, then cancels an active run or clears a terminal one.MITDestructive
- Create a model export job for a project or model with a format and optional settings. Requires confirm_cost=true; format validates upfront, but compatibility failures show only after the job runs.MIT
- Retrieve the status of an auto-annotation run for a dataset. Returns active job progress, last run results, or errors when the run fails.MIT
- Retrieve metrics for a specific deployment by owner/slug or bare slug, with configurable time range and optional compact sparkline for 24-hour request stats.MIT
- Report a model's training status and progress for public or private Ultralytics projects, showing epoch counters, progress percentage, ETA, and optional trainResults history.MIT
- Report a model's best training epoch and metrics for private or public projects, separating best-epoch values from top-level reported metrics and flagging mismatches.MIT
- Cancel an active export job for a model; checks status first, but a race condition can delete the finished artifact irreversibly instead.MITDestructive
- Run inference on a local image or video through a deployment's serving endpoint by owner/deployment or slug, returning images and metadata verbatim.MIT
- List models within a project by slug, owner/slug, or ul:// URI; for model quality, use model_metrics.MIT