$GRAZE CA: only trust the address shown on usegraze.world and posted by @usegraze. Claude, OpenAI and Kimi credits · 28 Sep
Graze docs

Quickstart

Point any OpenAI-compatible client at Graze and use graze/auto as the model. That is the whole integration.

1. Get a key

Self-serve signup is being built. Until it ships, message @usegraze and we will create an agent key with starter credit. Graze shows the key once; store it like a password.

2. Make a call

curl https://graze-api-production.up.railway.app/v1/chat/completions \
  -H "Authorization: Bearer $GRAZE_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "graze/auto", "max_tokens": 200,
       "messages": [{"role": "user", "content": "Summarise this token launch in one line."}]}'
from openai import OpenAI

client = OpenAI(base_url="https://graze-api-production.up.railway.app/v1", api_key="graze_your_key")
reply = client.chat.completions.create(
    model="graze/auto",
    max_tokens=200,
    messages=[{"role": "user", "content": "Summarise this token launch in one line."}],
)
print(reply.choices[0].message.content)

3. Read the receipt

Every response carries these headers:

Header Meaning
X-Graze-Model The model that answered
X-Graze-Attempts How many models were tried
X-Graze-Cost-USD What this call cost you, fee included
X-Graze-Saved-USD Saved against Claude Sonnet 4.6 list price

4. Check your balance

curl https://graze-api-production.up.railway.app/v1/balance -H "Authorization: Bearer $GRAZE_KEY"