Models

The model catalog is publicly readable — no API key required. Use it to discover available models, check DA pricing, and read capability flags before deciding which model to call.

List models#

The OpenAI SDK's client.models.list() returns a basic list. For richer filtering, call the catalog endpoint directly:

import os
import httpx

# No auth needed for the public catalog
resp = httpx.get(
    "https://opendunes.com/api/models",
    params={
        "provider": "anthropic",
        "capability": "vision",
        "sort": "popularity",
    },
)
models = resp.json()
for m in models["data"]:
    print(m["slug"], m["pricing"]["input_da_per_million"], "DA/M input tokens")

Filter parameters#

Retrieve a single model#

import httpx

slug = "meta-llama/llama-4-maverick"
resp = httpx.get(f"https://opendunes.com/api/models/{slug}")
model = resp.json()
print(model["name"])
print(model["pricing"])   # input/output DA per million tokens
print(model["context_length"])

Facets and stats#

GET /api/models/facets returns aggregated filter counts (providers, capabilities, modalities) — useful for building search UIs.

GET /api/models/stats returns catalog-wide counts and the list of available providers.

Browse in the UI#

You can also explore the full catalog at /models — search, filter by capability, and see per-conversation DA cost estimates.