APIGOTO

Inspiration

What it does

How I built it

Challenges I ran into

Accomplishments that I'm proud of

What I learned

What's next for APIGOTO

Disclosure: I work on the APIGOTO team. APIGOTO is an independent LLM API gateway and is not affiliated with OpenAI, Anthropic or Google.

Why we built it Apps that call more than one model provider usually end up with several SDKs, several keys and several sets of error handling. APIGOTO puts one OpenAI-compatible endpoint in front of Claude, GPT and Gemini, so existing OpenAI SDK code only needs a different base_url and key.

What it does

One OpenAI-compatible base_url: https://api.apigoto.com/v1
Automatic failover, capped at 3 retries, so a failing upstream does not turn into a retry storm
Per-request logs, so you can see what happened to each call
BYOK: bring your own provider keys if you prefer
Pay-per-use, no subscription

Minimal example (Python, OpenAI SDK)

from openai import OpenAI
client = OpenAI(base_url="https://api.apigoto.com/v1", api_key="YOUR_APIGOTO_KEY")
r = client.chat.completions.create(model="<model name from the APIGOTO model list>", messages=[{"role":"user","content":"Hello"}])
print(r.choices[0].message.content)

Open questions we'd like feedback on