A Groq API alternative with one key for 50+ models

What developers use it for

• Speed-sensitive inference on the open models Groq hosts
• Real-time and voice-style apps

Things to consider

• The catalogue is the set of models Groq hosts, which does not include Anthropic's Claude or OpenAI's closed GPT models.
• Check Groq's pricing page for current per-token rates.

Where Somnus fits

If you need closed frontier models such as Claude or GPT next to open models, Somnus gives you both from one balance.

Every Somnus top-up is credited at 2× or more, and usage is billed at each model's list price, so a dollar of credit buys at least two dollars of usage.

Switching from Groq: change two settings

from openai import OpenAI
import os

client = OpenAI(
    api_key=os.environ["SOMNUS_API_KEY"],
    base_url="https://somnus.world/api/public/gateway/v1",
)

resp = client.chat.completions.create(
    model="claude-sonnet-5",
    messages=[
      {"role": "user", "content": "Hello"}
    ],
)
print(resp.choices[0].message.content)

If you already use Groq's OpenAI-compatible endpoint, only base_url, the key and the model name change.

Questions

Related