Portkey alternative.
One endpoint. Every modality. Half the config.

from portkey_ai import Portkey

client = Portkey(
    api_key="pk-...",
    virtual_key="openai-...",
)

r = client.chat.completions.create(
    model="gpt-5.4",
    messages=[{"role": "user", "content": "Hello"}],
)

# pip install aigateway-py openai
# aigateway-py: sub-accounts, evals, replays, jobs, webhook verify.
# openai SDK: chat/embeddings/images/audio — drop-in compat per our SDK's own guidance.
from openai import OpenAI

client = OpenAI(
    base_url="https://api.aigateway.sh/v1",
    api_key="sk-aig-...",
)

r = client.chat.completions.create(
    model="openai/gpt-5.4",
    messages=[{"role": "user", "content": "Hello"}],
)

• 1150+ models across every modality, not 40 text-heavy.
• No seats, no subscription — you pay for usage only.
• Sub-account API mints end-user keys with spend caps and isolated analytics.
• Eval-driven routing: upload a dataset, we pick the cheapest model that passes.
• Cached requests get a 50% discount on the uncached cost (not 100%).

• Portkey ships a Python SDK with richer prompt-management primitives out of the box; we publish equivalents on our Enterprise tier.