FAQ

Common questions and answers about using the Regolo API.

Overview

This page addresses frequently asked questions about API parameters, model limits, and troubleshooting common issues.

Questions

How do I disable model fallbacks?

By default, if the requested model is unavailable, the API automatically falls back to a compatible alternative. To disable this behavior and receive an error instead, set disable_fallbacks to true.

Important

When disable_fallbacks is true, requests will fail with an error if the selected model is unavailable. Use this when you need guaranteed model consistency.

curl -L -X POST 'https://api.regolo.ai/v1/chat/completions' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -d '{
    "messages": [
      {
        "role": "user",
        "content": "List 5 important events in the 19th century"
      }
    ],
    "model": "qwen3.5-122b",
    "disable_fallbacks": true
  }'

What is the maximum value for `max_tokens`?

The maximum value for max_tokens varies by model. Setting a value higher than the model's limit will automatically bcrash the request.

curl -L -X POST 'https://api.regolo.ai/v1/chat/completions' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -d '{
    "messages": [
      {
        "role": "user",
        "content": "Write a long story"
      }
    ],
    "model": "qwen3.5-122b",
    "max_tokens": 120000
  }'

Cost Consideration

Higher max_tokens values increase latency and token usage. Set only what you need for your use case.

Can I generate multiple completions at once?

Yes, use the n parameter to generate multiple response variations in a single request.

curl -L -X POST 'https://api.regolo.ai/v1/chat/completions' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -d '{
    "messages": [
      {
        "role": "user",
        "content": "Suggest a name for a new API product"
      }
    ],
    "model": "qwen3.5-122b",
    "n": 3
  }'

The response will contain 3 different choices in the choices array.

How do I choose between creativity and accuracy?

Use the temperature parameter to control the balance:

Temperature	Use Case
0.0 - 0.3	Factual, accurate responses
0.4 - 0.7	Balanced (default)
0.8 - 1.0	Creative, varied outputs

# For factual answers
curl -L -X POST 'https://api.regolo.ai/v1/chat/completions' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -d '{
    "messages": [
      {
        "role": "user",
        "content": "What year did WW2 end?"
      }
    ],
    "model": "qwen3.5-122b",
    "temperature": 0.2
  }'

How do I reduce repetition in model outputs?

Use frequency_penalty to reduce token repetition and presence_penalty to encourage new topics.

# Reduce repetition
curl -L -X POST 'https://api.regolo.ai/v1/chat/completions' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -d '{
    "messages": [
      {
        "role": "user",
        "content": "List 10 different programming languages"
      }
    ],
    "model": "qwen3.5-122b",
    "frequency_penalty": 0.5,
    "presence_penalty": 0.4
  }'

Parameter	Range	Effect
frequency_penalty	-2.0 to 2.0	Reduces repeated tokens
presence_penalty	-2.0 to 2.0	Encourages new topics

FAQ