Models¶

Argo Proxy dynamically fetches the model list from the upstream ARGO API at startup. Rather than maintaining a static list of models (which quickly becomes outdated), this page explains the naming scheme used by Argo Proxy and how to query available models at runtime.

Querying Available Models¶

List Models¶

Send a GET request to the /v1/models endpoint to retrieve all currently available models:

curl http://localhost:44497/v1/models

The response follows the OpenAI-compatible format:

{
    "object": "list",
    "data": [
        {
            "id": "argo:gpt-4o",
            "internal_name": "gpt4o",
            "object": "model",
            "created": 1700000000,
            "owned_by": "openai"
        },
        ...
    ]
}

Each entry contains:

id — the Argo Proxy alias you use in API requests (e.g. argo:gpt-4o)
internal_name — the upstream ARGO internal model identifier (e.g. gpt4o)
owned_by — the model provider family (openai, anthropic, google, or unknown)

Refresh Model List¶

If new models are added upstream, you can reload the model list without restarting:

curl -X POST http://localhost:44497/refresh

See Endpoints — /refresh for details on the response format.

Model Naming Scheme¶

All Argo Proxy model names use the argo: prefix followed by a human-readable, OpenAI-style name. The naming rules vary by model family.

OpenAI Models¶

Standard GPT models use the format argo:gpt-{version}:

Pattern	Example
`argo:gpt-{version}`	`argo:gpt-4`, `argo:gpt-4-turbo`, `argo:gpt-4o`
`argo:gpt-{version}-{variant}`	`argo:gpt-3.5-turbo-16k`, `argo:gpt-4.1-mini`

OpenAI reasoning models (o-series) have two equivalent aliases:

Pattern	Example
`argo:gpt-{o-model}`	`argo:gpt-o3-mini`
`argo:{o-model}`	`argo:o3-mini`

Both forms resolve to the same upstream model. Use whichever you prefer.

Anthropic Claude Models¶

Claude models have two equivalent aliases with different ordering:

Pattern	Example
`argo:claude-{codename}-{generation}`	`argo:claude-sonnet-4`, `argo:claude-opus-4`
`argo:claude-{generation}-{codename}`	`argo:claude-4-sonnet`, `argo:claude-4-opus`

For versioned models, a version suffix is appended:

Pattern	Example
`argo:claude-{codename}-{generation}-{version}`	`argo:claude-sonnet-3.5-v2`
`argo:claude-{generation}-{codename}-{version}`	`argo:claude-3.5-sonnet-v2`

Google Gemini Models¶

Gemini models use the format argo:gemini-{version}-{variant}:

Pattern	Example
`argo:gemini-{version}-{variant}`	`argo:gemini-2.5-pro`, `argo:gemini-2.5-flash`

Embedding Models¶

Embedding models follow OpenAI's naming convention:

Pattern	Example
`argo:text-embedding-{name}`	`argo:text-embedding-ada-002`, `argo:text-embedding-3-small`

Flexible Model Name Resolution¶

Argo Proxy is lenient when resolving model names. The following variations are all accepted:

Prefix: argo:gpt-4o or just gpt-4o (the argo: prefix is optional)
Separator: argo:gpt-4o or argo/gpt-4o (slash works as well)
Case: argo:GPT-4o or argo:gpt-4o (case-insensitive)

If a model name cannot be resolved, Argo Proxy falls back to gpt4o for chat models and v3small for embedding models.

Why No Static Model List?¶

The upstream ARGO API evolves over time — models are added, retired, or renamed. Argo Proxy fetches the model list dynamically at startup and generates aliases automatically based on the naming rules above. This means:

New models are available immediately after an upstream update (just call /refresh or restart).
Documentation stays accurate without manual updates.
You always have the ground truth via the /v1/models endpoint.