[DOCS]
Quickstart
Blaze Inference is an OpenAI-compatible API for frontier open-weight models. If your code works with OpenAI, it works with Blaze: swap the base URL and you are done.
1 · Get a key
Sign in with your email (a 6-digit code, no password), then hit Create API key on the dashboard. The key is shown once; store it as BLAZE_API_KEY. You get 100 free requests a day; only successful responses count.
2 · Make a request
CURL
curl https://blazeinference.com/v1/chat/completions \
-H "Authorization: Bearer $BLAZE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "z-ai/glm-5.2",
"messages": [{"role": "user", "content": "Say hi in five words."}]
}'JAVASCRIPT · openai sdk
import OpenAI from "openai";
const blaze = new OpenAI({
apiKey: process.env.BLAZE_API_KEY,
baseURL: "https://blazeinference.com/v1",
});
const res = await blaze.chat.completions.create({
model: "deepseek-ai/deepseek-v4-pro",
messages: [{ role: "user", content: "Say hi in five words." }],
});
console.log(res.choices[0].message.content);PYTHON · openai sdk
import os
from openai import OpenAI
blaze = OpenAI(
api_key=os.environ["BLAZE_API_KEY"],
base_url="https://blazeinference.com/v1",
)
res = blaze.chat.completions.create(
model="deepseek-ai/deepseek-v4-pro",
messages=[{"role": "user", "content": "Say hi in five words."}],
)
print(res.choices[0].message.content)3 · Point your harness at it
Works out of the box with anything that accepts an OpenAI-compatible endpoint: opencode, Claude Code, Cline, Roo Code, Zed, aider. Set the base URL to https://blazeinference.com/v1 and paste your key. No other configuration is required.
Next
Chat completions for streaming and every supported parameter · Models for what's live · Errors for the full error-code list.