Choosing an AI API for classification and structured output

What classification and structured output needs

• JSON output your code can parse
• Very low cost per item
• Consistent labels

Models that fit

Somnus column uses the minimum 2× credit; larger top-ups go further.

Cost example

1,000 requests of about 1,500 input and 400 output tokens on DeepSeek V4 Flash cost about $0.59 at list price — roughly $0.30 with Somnus credit.

Tips

• Request JSON output and validate it with a schema.
• List the allowed labels explicitly.
• Test on a sample of 100 items before running thousands.

Common mistakes

• Free-text answers that break your parser.
• Paying flagship prices for a yes/no decision.

Example

from openai import OpenAI
import os

client = OpenAI(
    api_key=os.environ["SOMNUS_API_KEY"],
    base_url="https://gateway-production-c837.up.railway.app/v1",
)

resp = client.chat.completions.create(
    model="claude-sonnet-5",
    messages=[
      {"role": "system", "content": "Reply only with JSON."},
      {"role": "user", "content": "Classify this message as billing, bug, or feature_request: 'I was charged twice.'"}
    ],
    response_format={"type": "json_object"},
)
print(resp.choices[0].message.content)

Questions

Related