> ## 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.

# Routing and versions

> Control target weights and parameters, and isolate changes before production rollout.

Routing defines which model serves a project version. Every target contains a model reference, non-negative weight, and optional parameter overrides.

## Weights are relative

Weights don't need to add to 100. A configuration with weights `3` and `1` sends approximately 75% and 25%. A zero-weight target stays available in the configuration but receives no new weighted traffic.

Use zero weight to preserve the previous model for visibility and rollback after a full promotion.

## Sticky routing

When a request supplies `gen_ai.conversation.id` in `metadata`, the proxy deterministically keeps that conversation on one target for the routing revision. Without a conversation ID, the proxy selects a target independently for each request.

## Target parameters

To add values such as temperature, reasoning effort, response format, or provider-specific controls, expand **Parameters** under a model. Numbers, booleans, arrays, and objects are parsed from JSON. Other text remains a string.

The stored override wins over the same field supplied by the caller.

A model override, which sends a `provider/model` value in `model` instead of `"router"`, bypasses the targets. It uses the caller's parameters and the provider's defaults, not the targets' parameters. The proxy traces, samples, and caches it in the selected version like routed traffic.

## Save to an existing or new version

Open the project's **Routing** page. Add or edit targets, choose the destination in the version bar, and select **Save routing**. The version bar is part of the same save action:

* Choose an existing version to replace its desired routing.
* Choose the new-version option to create a new version. The version is created only when you select **Save routing**. To call the new version, set `forge.project_version` in the request `metadata` to its number, such as `"2"`. Requests without `forge.project_version` use version 1.

Each save publishes a new routing revision: a numbered configuration for that version that includes its targets, weights, parameter overrides, and trace-sampling rate. Publishing is asynchronous, so a saved configuration serves traffic only after its deployment status becomes **Applied** at that revision.

> [!WARNING]
>
> If you delete a version, callers pinned to it receive `404` responses. You can't delete version 1.

> [accordion] API: Replace routing for a project version (PUT /projects/\{alias}/versions/\{version}/routing)
>
> Routing updates replace the whole configuration, so the request must include `trace_sampling_rate` as well as every target. To keep the existing sampling rate, pass it in the request. If you want the previous model available for rollback, keep it at weight `0`:
>
> ```bash theme={"system"}
> curl --request PUT \
>   --url "https://forge.coreweave.com/api/distillation/v1/projects/ticket-classifier/versions/2/routing" \
>   --header "Authorization: Bearer $WANDB_API_KEY" \
>   --header "Forge-Entity: your-team" \
>   --header "Content-Type: application/json" \
>   --data '{
>     "trace_sampling_rate": 1,
>     "targets": [
>       {"model_ref": "openai/gpt-5.6-sol", "weight": 0},
>       {"model_ref": "wandb-inference/your-fine-tuned-model", "weight": 1}
>     ]
>   }'
> ```
>
> Poll the returned routing state until deployment is `applied` at the current revision. See [Replace routing for a project version](https://docs.coreweave.com/model-distillation/reference/management/routing/replace-routing-for-a-project-version.md).
