---
title: "Discover account-qualified managed models and prices"
method: GET
path: "/v1/managed-models/catalog"
tags: ["Managed models"]
---

# Discover account-qualified managed models and prices

`GET /v1/managed-models/catalog`

## Request

- Base URL: `/` (relative, and the document was published on a file host, not at the API’s origin)
- URL: `/v1/managed-models/catalog`
- Auth: HTTP bearer

Requires a Nodus API key or console session. Returns only qualified Messages models explicitly granted to this account. Accounts without configured access receive enabled false and an empty models array. Discovery does not start compute or make a model call. Prices are customer USD micros per million tokens and include the service markup. Managed model access does not install agent harnesses or inject credentials into hosted workers.

## Response `200`

Qualified catalog, including an empty catalog when access is not enabled.

- ManagedModelCatalog
  - `enabled` boolean, required: True when at least one model is granted to this account.
  - `models` ManagedModelPublic[], required: Qualified models sorted by ID. Empty for accounts without a configured grant.
    - `id` string, required: Model ID returned by this account’s managed model catalog.
    - `label` string, required
    - `protocol` 'messages' | 'chat_completions', required
    - `price_version` string, required
    - `rates` ManagedModelRates, required: Published customer rates include the 12.5 percent service markup and round upward for display. One USD micro is 0.000001 USD. Each token belongs to one billing bucket. Actual charges apply the markup to the aggregate selected model supplier cost and round upward once. When rates_are_maximum is true, rates are the maximum across eligible models and the actual charge may be lower.
      - `input_micros_per_million` integer, required: Customer price for uncached input, in USD micros per million tokens.
      - `output_micros_per_million` integer, required: Customer price for output including reasoning, in USD micros per million tokens.
      - `cache_read_micros_per_million` integer: Customer price for cached input reads. Omission means this cache bucket is not qualified, not free.
      - `cache_write_5m_micros_per_million` integer: Customer price for five-minute cache creation. Omission means this bucket is not qualified.
      - `cache_write_1h_micros_per_million` integer: Customer price for one-hour cache creation. Omission means this bucket is not qualified.
    - `rates_are_maximum` boolean: When true, published rates are maximum customer rates across eligible models. Actual charges use the selected model supplier cost plus 12.5 percent.
    - `max_output_tokens` integer, required
    - `streaming` false, required
  - `billing` 'nodus_credits', required
  - `currency` 'USD', required
  - `assistant_templates` ManagedAssistantTemplate[], required: Built-in assistants available on the qualified runtime. Empty until account, compute, model and exact image qualifications all pass.
    - `id` 'nodus:claude-assistant-v1' | 'nodus:claude-assistant-v2', required
    - `label` string, required
    - `protocol` 'messages', required
    - `model_ids` string[], required: Models qualified for this installed assistant.
    - `default_output_tokens_by_model` object: Default output allowance for each available catalog model.
    - `default_output_tokens` integer, required: Largest default among available models. Use default_output_tokens_by_model for a selected model.

## Other responses

- `default`: Authentication, request-rate or transient failure.

## Changes

- **2026-10-01** `526f94ab6f21`: 1 info
  - added the optional property `models/items/rates_are_maximum` to the response with the `200` status
- **2026-10-01** `24ef3f281240`: 1 breaking, 1 info
  - the `default_output_tokens` response property const value `1024` was removed for the status `200`
  - added the optional property `assistant_templates/items/default_output_tokens_by_model` to the response with the `200` status
- **2026-10-01** `50087398f13a`: 1 breaking, 2 warning
  - the `id` response property const value `nodus:claude-assistant-v1` was removed for the status `200`
  - added the new `nodus:claude-assistant-v1` enum value to the `assistant_templates/items/id` response property for the response status `200`
  - added the new `nodus:claude-assistant-v2` enum value to the `assistant_templates/items/id` response property for the response status `200`
- **2026-10-01** `ba84c74aeec8`: 1 info
  - endpoint added
- **2026-10-01** `544fef290a54`: 1 breaking
  - api path removed without deprecation

[Operation history](https://skmtc.dev/nodus-compute/apis/nodus-infrastructure-execution-api/changes/v1/managed-models/catalog/get.md)

---

[API](https://skmtc.dev/nodus-compute/apis/nodus-infrastructure-execution-api.md) - [All operations](https://skmtc.dev/nodus-compute/apis/nodus-infrastructure-execution-api/llms.txt) - [OpenAPI document](https://skmtc.dev/nodus-compute/apis/nodus-infrastructure-execution-api/revisions/a59210899d95?raw)