Every team building with AI ends up in the same place: a drawer full of API keys. One for OpenAI, one for Anthropic, one for Google, another for whatever model shipped last week. Each with its own SDK, its own billing portal, its own rate limits, its own idea of what a "message" is.
Troviale collapses that drawer into a single key.
One endpoint, every provider
Point your existing OpenAI SDK at Troviale, drop in one key, and call any model
by name — claude-opus-4-7, claude-sonnet-4-6, and more as we add them.
Switching models is a string change, not a migration.
from openai import OpenAI
client = OpenAI(
base_url="https://api.troviale.com/v1",
api_key="YOUR_TROVIALE_KEY",
)
client.chat.completions.create(
model="claude-sonnet-4-6",
messages=[{"role": "user", "content": "Hello"}],
)One wallet, usage-based
No per-provider subscriptions and no minimums. You top up a single balance and spend it across every model at transparent per-token pricing. Usage and spend land in one dashboard instead of five.
Pricing for every model — input and output, per million tokens — is public on the models page. No sign-up required to compare.
What's next
This is the foundation. We're expanding the catalog, adding deeper observability, and publishing guides for migrating off single-provider setups. Follow along on the changelog.
One key, every model.