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

# Models

`GET /v1/models`

Returns the list of models currently available on the network. Availability reflects the live worker pool: a model appears in this list only if at least one worker is currently hosting it.

***

## List models

```bash
curl https://api.usedecentral.org/v1/models \
  -H "Authorization: Bearer dcntrl_live_your_key_here"
```

### Response

```json
{
  "object": "list",
  "data": [
    {
      "id": "llama-3.3-70b",
      "object": "model",
      "created": 1750000000,
      "owned_by": "meta",
      "decentral": {
        "tier": "max",
        "credits_per_request": 40,
        "credits_per_1k_tokens": 20,
        "active_workers": 12,
        "median_latency_ms": 1840,
        "parameters": "70B",
        "context_window": 128000
      }
    },
    {
      "id": "qwen3-8b",
      "object": "model",
      "created": 1750000000,
      "owned_by": "qwen",
      "decentral": {
        "tier": "standard",
        "credits_per_request": 8,
        "credits_per_1k_tokens": 4,
        "active_workers": 47,
        "median_latency_ms": 420,
        "parameters": "8B",
        "context_window": 32768
      }
    }
  ]
}
```

The `decentral` object contains additional fields not present in the standard OpenAI response.

| Field                   | Description                                                |
| ----------------------- | ---------------------------------------------------------- |
| `tier`                  | Pricing tier: `lite`, `standard`, `pro`, or `max`          |
| `credits_per_request`   | Credit cost for completions up to \~500 output tokens      |
| `credits_per_1k_tokens` | Credit cost per 1,000 output tokens for longer completions |
| `active_workers`        | Number of workers currently hosting this model             |
| `median_latency_ms`     | Observed p50 time-to-first-token across the worker pool    |
| `parameters`            | Parameter count                                            |
| `context_window`        | Maximum context length in tokens                           |

***

## Available models

Models available at launch. The list grows as workers add new models and governance votes approve additions.

| Model ID        | Parameters | Context | Tier     | Credits/request |
| --------------- | ---------- | ------- | -------- | --------------- |
| `llama-3.3-70b` | 70B        | 128K    | Max      | 40              |
| `deepseek-r1`   | 70B        | 128K    | Max      | 40              |
| `llama-3.2-27b` | 27B        | 128K    | Pro      | 18              |
| `qwen3-14b`     | 14B        | 32K     | Pro      | 18              |
| `qwen3-8b`      | 8B         | 32K     | Standard | 8               |
| `mistral-7b`    | 7B         | 32K     | Standard | 8               |
| `llama-3.2-3b`  | 3B         | 128K    | Lite     | 2               |
| `qwen3-1.7b`    | 1.7B       | 32K     | Lite     | 2               |

Model IDs are stable. When an underlying model is superseded, the old ID remains valid until a governance vote retires it, with advance notice.

***

## Model availability

A model's `active_workers` count reflects the live state of the network. Models with more active workers have lower latency and better job acceptance rates.

If you request a model with zero active workers, the API returns a `503` with a `retry_after` suggestion.

During beta, the Decentral team maintains a first-party seed provider pool to ensure all launch models have consistent availability for early users.

***

## Retrieve a specific model

```bash
curl https://api.usedecentral.org/v1/models/qwen3-8b \
  -H "Authorization: Bearer dcntrl_live_your_key_here"
```

Returns a single model object in the same format as the list response. Useful for checking the current worker count and latency before submitting a job.


---

# 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 dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.usedecentral.org/api-reference/models.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

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.