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

# Write your manifest

The manifest is a JSON file that tells the network what your node offers: which models, how many requests at once, and at what price.

## A minimal manifest

```json
{
  "node": "example-node-01",
  "models": [
    { "id": "meta-llama/llama-3.3-70b", "slots": 32 }
  ]
}
```

That's enough to start. Only three things are required:

* `node`: a name for this node, shown in your reports.
* `models[].id`: the model id **exactly** as your engine's `/v1/models` lists it. Clients ask for the model by this id.
* `models[].slots`: the most requests for this model you'll serve at once. The network never sends more.

## A fuller example

```json
{
  "node": "example-node-01",
  "models": [
    {
      "id": "meta-llama/llama-3.3-70b",
      "quant": "fp8",
      "context_max": 32768,
      "slots": 32,
      "interactive_reserve": 4,
      "speed": { "single_tps": 20, "batch_tps": 250 }
    }
  ],
  "tiers": ["interactive", "batch"],
  "price": {
    "per_mtok_in": 0.05,
    "per_mtok_out": 0.40,
    "batch": { "per_mtok_in": 0.03, "per_mtok_out": 0.25 }
  }
}
```

Every field is described in the [manifest reference](/reference/manifest.md).

## Choosing `slots`

Set `slots` to the number of requests your engine handles well at the same time, not the most it accepts. Too high and every request gets slow; too low and you leave money on the table. A quick way to find it: run a load test against your engine and watch time to first token as concurrency goes up.

## Be honest

Everything in the manifest is a claim. The network checks it against what your node actually does, so describe the model and precision you really run.

## Changing it later

Edit the file while the agent runs. It notices within 5 seconds (or right away on `SIGHUP`), checks your engine serves every listed model, and sends the new manifest. Requests already running carry on. If the new file is invalid, the agent logs it and keeps the old one.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.podtorrent.com/running-a-node/manifest.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.