Regolo API is not working

[dfigos-2000]

Hi,
I wanted to use my API key from Regolo, and Omniroute cannot retrive models and test Regolo.

API unavailable — using local catalog
No new models to import — all models are already in the registry or custom models list

Also, I see that, even when is not working, the window that appears shows All models already imported with a "correct" symbol. I guess for these situations, should have a "wrong" or "faulty" symbol.

Here is a brief from Regolo docs:

Regolo provides an OpenAI-compatible API. Get your API key from Regolo, choose a model from the model catalog
, and call the Chat Completions endpoint.

Quick start

export REGOLO_API_KEY="your-api-key"

curl https://api.regolo.ai/v1/chat/completions \
  -H "Authorization: Bearer $REGOLO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-oss-120b",
    "messages": [
      {
        "role": "user",
        "content": "Explain recursion in one paragraph."
      }
    ]
  }'

Reasoning

For models supporting reasoning, add reasoning_effort:

curl https://api.regolo.ai/v1/chat/completions \
  -H "Authorization: Bearer $REGOLO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-oss-120b",
    "messages": [
      {
        "role": "user",
        "content": "Solve this problem step by step."
      }
    ],
    "reasoning_effort": "high"
  }'

The final answer is returned in choices[0].message.content; reasoning output, when provided, is available in choices[0].message.reasoning_content.

Python
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["REGOLO_API_KEY"],
    base_url="https://api.regolo.ai/v1",
)

response = client.chat.completions.create(
    model="gpt-oss-120b",
    messages=[
        {"role": "user", "content": "Hello from Regolo!"}
    ],
)

print(response.choices[0].message.content)

References

Model Catalog

API Documentation

Reasoning

[diegosouzapw]

Thanks @dfigos-2000, you found a real bug in our Regolo entry, not a problem with your key.

The built-in regolo provider has its endpoint set to the bare host https://api.regolo.ai (open-sse/config/providers/registry/regolo/index.ts), with no /v1/chat/completions path and no models URL. The default executor uses that value as the literal request URL, so chat requests go to the root of the host. Model discovery has nothing to call either, which is why you get "API unavailable — using local catalog". Its two built-in model ids (regolo-chat, regolo-fast) don't exist in Regolo's catalog either. It's tracked in #14996.

Workaround that works today: edit the Regolo connection, open Advanced, and set the Base URL override to:

https://api.regolo.ai/v1

OmniRoute turns that into https://api.regolo.ai/v1/chat/completions. Regolo accepts any model id through OmniRoute, so call the real ids directly, e.g. regolo/gpt-oss-120b. Model import may still come up empty until the fix lands, but chat doesn't depend on it.

You're also right about the dialog. "All models already imported" with a success check, shown when the API call actually failed, hides the failure. That's in the same issue.

[diegosouzapw]

@dfigos-2000 Fixed: #15000 points the Regolo provider at its real endpoints (https://api.regolo.ai/v1/chat/completions, with model discovery through /v1/models) and drops the two made-up model ids. It's merged into release/v3.8.51, so it's in the next :next image and in v3.8.51. Once you're on either, you can remove the Base URL override. The model import should then list Regolo's actual catalog.

The misleading "All models already imported" check shown when discovery fails wasn't part of that fix and hasn't been fixed yet.