> For the complete documentation index, see [llms.txt](https://docs.roboflow.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.roboflow.com/models/readme.md).

# Models

Train state-of-the-art computer vision models on your datasets - no ML expertise or GPU setup required. Roboflow manages the training infrastructure and hosts the resulting models for deployment.

## Choose & Train

<table data-view="cards"><thead><tr><th></th><th></th><th data-hidden data-card-cover data-type="image">Cover image</th><th data-hidden data-card-target data-type="content-ref"></th></tr></thead><tbody><tr><td><strong>Supported Models</strong></td><td>Every model you can train, upload, or deploy with Roboflow, and where each runs.</td><td><a href="/files/LWhDJOPhmQE6x7Rb0HLd">/files/LWhDJOPhmQE6x7Rb0HLd</a></td><td><a href="/pages/pzMXw9YN6MHLB8PihRgh">/pages/pzMXw9YN6MHLB8PihRgh</a></td></tr><tr><td><strong>Train a Model</strong></td><td>Start a training job from the dashboard, API, SDK, or CLI.</td><td><a href="/files/mPk0IsEpnecqcEeXuRre">/files/mPk0IsEpnecqcEeXuRre</a></td><td><a href="/pages/tG2aS2EaKtqtB8QjnhmI">/pages/tG2aS2EaKtqtB8QjnhmI</a></td></tr><tr><td><strong>Roboflow Instant</strong></td><td>A model trained automatically from your first approved annotations.</td><td><a href="/files/63EkjtP6q60Z0a15fc9I">/files/63EkjtP6q60Z0a15fc9I</a></td><td><a href="/pages/DMf5zHlWHkAyX5UxzVLG">/pages/DMf5zHlWHkAyX5UxzVLG</a></td></tr><tr><td><strong>Neural Architecture Search</strong></td><td>Discover and fine-tune novel architectures.</td><td><a href="/files/rdCm11yl9UNZeTgdhOqp">/files/rdCm11yl9UNZeTgdhOqp</a></td><td><a href="/pages/c7yT3Wd0tZKgE1QAVcbW">/pages/c7yT3Wd0tZKgE1QAVcbW</a></td></tr><tr><td><strong>Roboflow Rapid</strong></td><td>Build a working model from a small sample of data in minutes.</td><td><a href="/files/9A4jcQYN4CbCvmNzlr9y">/files/9A4jcQYN4CbCvmNzlr9y</a></td><td><a href="/pages/9AeIVz7E6PBtIOB0ZqpC">/pages/9AeIVz7E6PBtIOB0ZqpC</a></td></tr><tr><td><strong>Train from a Checkpoint</strong></td><td>Start training from an existing Roboflow Universe model.</td><td><a href="/files/PUuo0L7NmrJz0HNSG3fg">/files/PUuo0L7NmrJz0HNSG3fg</a></td><td><a href="/pages/mcmEc0tgn3qfXRi7niQ9">/pages/mcmEc0tgn3qfXRi7niQ9</a></td></tr></tbody></table>

You can also train from a third-party checkpoint on [Azure Vision](/models/train/pro-third-party-training-integrations.md) or [Google Cloud](/models/train/train-from-google-cloud.md).

## Evaluate & Understand

<table data-view="cards"><thead><tr><th></th><th></th><th data-hidden data-card-cover data-type="image">Cover image</th><th data-hidden data-card-target data-type="content-ref"></th></tr></thead><tbody><tr><td><strong>Evaluate Trained Models</strong></td><td>Evaluation runs, mAP, and confusion matrices.</td><td><a href="/files/uKBCof7Qq5NIVZuKe3hC">/files/uKBCof7Qq5NIVZuKe3hC</a></td><td><a href="/pages/24kFbXI2alnqt5mRyH08">/pages/24kFbXI2alnqt5mRyH08</a></td></tr><tr><td><strong>Training Results</strong></td><td>Review metrics and graphs from a completed training job.</td><td><a href="/files/YhIMs8YIy3yJ1jwaKdjX">/files/YhIMs8YIy3yJ1jwaKdjX</a></td><td><a href="/pages/WXr5TF1z44hlEBGeBhBY">/pages/WXr5TF1z44hlEBGeBhBY</a></td></tr><tr><td><strong>View Trained Models</strong></td><td>Browse the models trained in your workspace.</td><td><a href="/files/Hb1mrPN2qpFAR28i7l8N">/files/Hb1mrPN2qpFAR28i7l8N</a></td><td><a href="/pages/DW1EbDP4ADBNWrpcuYKw">/pages/DW1EbDP4ADBNWrpcuYKw</a></td></tr><tr><td><strong>Training Resolutions</strong></td><td>See the input resolution used by each model type.</td><td><a href="/files/hZJQEpna7wsi0uUuJvru">/files/hZJQEpna7wsi0uUuJvru</a></td><td><a href="/pages/3Oi0qU9nx4vCG7NiHCaE">/pages/3Oi0qU9nx4vCG7NiHCaE</a></td></tr><tr><td><strong>Versions, Trainings, and Models</strong></td><td>How datasets, training jobs, and models relate.</td><td><a href="/files/NcsPA4o8OnQfGQ5q2oFy">/files/NcsPA4o8OnQfGQ5q2oFy</a></td><td><a href="/pages/IboE7UclItZfkfqfnjWP">/pages/IboE7UclItZfkfqfnjWP</a></td></tr><tr><td><strong>Model IDs</strong></td><td>How models are identified when you deploy or run inference.</td><td><a href="/files/kbQODR1p16dL4XDulM1e">/files/kbQODR1p16dL4XDulM1e</a></td><td><a href="/pages/kZ09PfT5MbQcc2oxARNK">/pages/kZ09PfT5MbQcc2oxARNK</a></td></tr></tbody></table>

Trained models deploy through the [Serverless API](https://docs.roboflow.com/deployment/roboflow-cloud/serverless-api), [Dedicated Deployments](https://docs.roboflow.com/deployment/roboflow-cloud/dedicated-deployments), or [self-hosted Inference](https://docs.roboflow.com/deployment/self-hosted/self-hosted).