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

# Quick start

> Connect a project, collect data, train candidate models, evaluate them, and deploy a winner.

This guide shows the complete path from an existing AI feature to a specialized model running in production. It's for developers who already have an application that calls a model through an OpenAI client and want to replace that model without changing the application each time.

You connect your application to CoreWeave Model Distillation, collect real traffic into a dataset, fine-tune and evaluate candidate models, and then route production traffic to the winner. By the end, your application selects a stable project through the proxy, and a candidate you chose for its balance of quality, speed, and cost serves the share of traffic you assign to it.

> [!NOTE]
>
> Before you begin, your organization must have [private-preview access to Model Distillation](https://docs.coreweave.com/model-distillation/overview.md#availability).

## Step 1: Sign in to Forge

In Forge, choose **Post-Training > Model Distillation**. Open [Model Distillation](https://forge.coreweave.com/post-train/distillation), sign in with your Forge API key, and choose the team that owns the application.

> [!NOTE]
>
> We recommend a team service-account key for a production application. It keeps your application's availability from depending on one employee's account.

## Step 2: Register your current model provider

Open **Providers** and connect the service that runs your current model. This lets Model Distillation keep serving the same model while you collect data and train alternatives.

If you use the built-in `wandb-inference` provider, you can skip this step.

The provider must expose an OpenAI-compatible Chat Completions endpoint for routed application traffic. Enter the exact API base URL, including any required path prefix such as `/v1`. The proxy appends `/chat/completions`. For UI relabeling and evaluations, the provider named `openai` uses OpenAI's Responses API instead.

After you save the provider, if the model ID isn't already available, add or enable it. Wait until the provider deployment is **Applied** at the current revision before continuing.

## Step 3: Create a project

From **Projects**, select **Create or import a project**. Create a new Weights & Biases project or import one you already have, give it a project alias such as `ticket-classifier`, and select the model that serves it today. The project alias defaults to the project name and identifies the project in Studio URLs and management API paths.

Your application selects the project with its `OpenAI-Project` value, `[ENTITY]/[PROJECT]`, such as `your-team/ticket-classifier`. `[PROJECT]` is the Weights & Biases project name, not the project alias. Studio shows this value in the project header. The value stays the same as you train and deploy new implementations.

> [accordion] API: Create a project (PUT /projects/\{alias})
>
> The project alias is the `alias` in the URL. Set the Weights & Biases project name in `project`, choose `create` or `import` in `project_mode`, and set the current model in `targets`:
>
> ```bash theme={"system"}
> curl --request PUT \
>   --url "https://forge.coreweave.com/api/distillation/v1/projects/ticket-classifier" \
>   --header "Authorization: Bearer $WANDB_API_KEY" \
>   --header "Forge-Entity: your-team" \
>   --header "Content-Type: application/json" \
>   --data '{
>     "project": "ticket-classifier",
>     "project_mode": "create",
>     "targets": [
>       {"model_ref": "openai/gpt-5.6-sol", "weight": 1}
>     ]
>   }'
> ```
>
> For the complete request and response, see [Create a Model Distillation project](https://docs.coreweave.com/model-distillation/reference/management/projects/create-a-model-distillation-project.md).

Routing is published asynchronously. Open the project's **Routing** page and wait until the deployment is **Applied** at the current routing revision before sending application traffic.

## Step 4: Connect your application

Keep your normal OpenAI client, but change its base URL, project, and model:

```python theme={"system"}
from openai import OpenAI

client = OpenAI(
    base_url="https://proxy.training.coreweave.com/v1",
    api_key=wandb_api_key,
    project="your-team/ticket-classifier",  # sent as OpenAI-Project
)

response = client.chat.completions.create(
    model="router",  # the project's weighted routing
    messages=messages,
    metadata={"gen_ai.conversation.id": conversation_id},
)
```

Model Distillation now collects useful examples from normal application traffic. For Python, JavaScript, and `curl` examples, see [Connect your application](https://docs.coreweave.com/model-distillation/get-started/connect-your-app.md).

## Step 5: Create a dataset

After the project has representative traffic, create a dataset. Review the examples and remove poor ones. If you want the new model to learn better behavior than the current model provides, improve the answers.

Follow the [Datasets Quick Start](https://docs.coreweave.com/model-distillation/studio/datasets-quickstart.md) to create and inspect your first dataset.

## Step 6: Fine-tune several models

Train one or more candidate models from the dataset. Trying several candidates gives you a better chance of finding the right balance of quality, speed, and cost.

Follow the [Fine-tuning Quick Start](https://docs.coreweave.com/model-distillation/studio/fine-tuning-quickstart.md) to start a training job.

## Step 7: Evaluate the candidates

Compare every candidate with the answers you want to preserve. Review the overall result and individual examples before choosing a winner.

Follow the [Evaluations Quick Start](https://docs.coreweave.com/model-distillation/studio/evaluations-quickstart.md) to run your first comparison.

## Step 8: Deploy the winner

Add the winner to project routing. You can start with a small share of traffic, increase it gradually, and keep the original model ready for rollback.

Your application keeps sending the same `OpenAI-Project` value with `model: "router"`, and the new model now serves the share of traffic you assigned to it.

> [card] [Train and deploy your first model](https://docs.coreweave.com/model-distillation/get-started/first-model.md)
>
> Follow the complete manual workflow with more detail about each decision.
