> ## Documentation Index
> Fetch the complete documentation index at: https://docs.primeintellect.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Model Factory Overview

> Train, evaluate, deploy and improve your own models on one platform

Model Factory is where you build your own models on Prime Intellect. You train a model, evaluate it, deploy it, and use what it does in production to train the next version. Each step runs on the same platform, so a checkpoint from one step is ready for the next.

Model Factory is built on our open-source libraries: [prime-rl](https://docs.primeintellect.ai/prime-rl/overview.md) for training and [verifiers](https://docs.primeintellect.ai/verifiers/overview.md) for environments and evaluations.

## The loop

| Step | What you do | Where |
| - | - | - |
| **Train** | Run reinforcement learning, supervised fine-tuning or pretraining on an open-weights model. | [Hosted Training](https://docs.primeintellect.ai/hosted-training/getting-started.md) |
| **Evaluate** | Measure a model on environments from the Environments Hub, or on your own. | [Hosted Evaluations](https://docs.primeintellect.ai/tutorials-environments/hosted-evaluations.md) |
| **Deploy** | Serve a model as a dedicated inference endpoint. | [Dedicated Endpoints](https://app.primeintellect.ai/dashboard/inference/dedicated) |
| **Improve** | Capture traces from your runs and endpoints, find where the model fails, and feed that into the next run. | [Prime Traces](https://docs.primeintellect.ai/traces/overview.md) |

## Two ways to run it

| | On-demand | Dedicated cluster |
| - | - | - |
| **What it is** | You submit a config and we run it on GPUs we operate. | A cluster reserved for your workloads alone. |
| **Access** | Self-serve | [Contact us](mailto:contact@primeintellect.ai) |
| **Billing** | [Per GPU-hour](https://docs.primeintellect.ai/hosted-training/billing.md), prorated to the second | Reserved capacity |
| **Use it when** | You want to start a run without reserving anything. | You need control over the stack: Slurm and Kubernetes access, or a pinned build of the training framework. |
| **Start here** | [Getting Started](https://docs.primeintellect.ai/hosted-training/getting-started.md) | [Model Factory Clusters](https://docs.primeintellect.ai/model-factory/clusters.md) |

To run the same training framework on your own hardware, use [prime-rl](https://docs.primeintellect.ai/prime-rl/overview.md) directly.

## Train

Hosted Training supports three kinds of run. Each one is dispatched with `prime train` and a TOML config.

> [card] [Reinforcement learning](https://docs.primeintellect.ai/hosted-training/hosted-rl.md)
>
> Full-parameter RL against a verifiers environment.

> [card] [Supervised fine-tuning](https://docs.primeintellect.ai/hosted-training/hosted-sft.md)
>
> Train on examples of the behavior you want.

> [card] [Pretraining](https://docs.primeintellect.ai/hosted-training/hosted-pretrain.md)
>
> Pretrain a model from scratch on your own corpus.

Run `prime train models` to see which models start warm and what each GPU type costs. See [Models](https://docs.primeintellect.ai/hosted-training/models.md) for details, and [Volumes](https://docs.primeintellect.ai/hosted-training/volumes.md) for how checkpoints are stored.

## Evaluate

Run an environment against a model from the dashboard or with `prime eval run --hosted`. Results appear on the [Evaluations page](https://app.primeintellect.ai/dashboard/evaluations). See [Evaluating environments](https://docs.primeintellect.ai/tutorials-environments/evaluating.md) for the local and hosted workflows.

## Deploy

Create a dedicated inference endpoint from the [Dedicated Endpoints page](https://app.primeintellect.ai/dashboard/inference/dedicated). An endpoint serves a model through an OpenAI-compatible API on GPUs reserved for it. For shared, pay-per-token models, see [Prime Inference](https://docs.primeintellect.ai/inference/overview.md).

## Improve

[Prime Traces](https://docs.primeintellect.ai/traces/overview.md) stores the traces from your training runs, evaluations and endpoints. Use them to see what the model did on each task and where it failed. A recurring failure can become an evaluation or a training task in your next run.

Trace uploads incur ingestion and storage charges. See [Pricing and settings](https://docs.primeintellect.ai/traces/pricing-and-settings.md).

## Next steps

> [card] [Getting Started](https://docs.primeintellect.ai/hosted-training/getting-started.md)
>
> Launch your first on-demand training run.

> [card] [End-to-End Run](https://docs.primeintellect.ai/hosted-training/end-to-end-run.md)
>
> Full walkthrough with baseline evals, model selection and result review.

> [card] [Model Factory Clusters](https://docs.primeintellect.ai/model-factory/clusters.md)
>
> Dedicated clusters running training, inference, Slurm and Kubernetes.

> [card] [Platform Overview](https://docs.primeintellect.ai/platform-overview.md)
>
> How environments, training, evaluation and inference fit together.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.