AI Code Generation API - Runcrate

Use purpose-built coding models to generate functions, review pull requests, and fix bugs programmatically. These models are trained specifically for code and outperform general-purpose models on programming tasks.

Recommended code models

Model	Strengths
`Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo`	480B MoE, fast inference, strong across all languages
`deepseek-ai/DeepSeek-V3.2`	Excellent reasoning, good for complex refactors

Generate a function

curl https://api.runcrate.ai/v1/chat/completions \
  -H "Authorization: Bearer rc_live_YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo",
    "messages": [
      {"role": "system", "content": "You are a senior software engineer. Write clean, production-ready code with error handling. Return only the code, no explanation."},
      {"role": "user", "content": "Write a Python function that retries an HTTP request with exponential backoff. Accept a URL, max retries (default 3), and base delay (default 1.0)."}
    ],
    "max_tokens": 1024
  }'

from openai import OpenAI

client = OpenAI(
    base_url="https://api.runcrate.ai/v1",
    api_key="rc_live_YOUR_API_KEY",
)

response = client.chat.completions.create(
    model="Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo",
    messages=[
        {"role": "system", "content": "You are a senior software engineer. Write clean, production-ready code with error handling. Return only the code, no explanation."},
        {"role": "user", "content": "Write a Python function that retries an HTTP request with exponential backoff. Accept a URL, max retries (default 3), and base delay (default 1.0)."},
    ],
    max_tokens=1024,
)
print(response.choices[0].message.content)

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.runcrate.ai/v1",
  apiKey: "rc_live_YOUR_API_KEY",
});

const response = await client.chat.completions.create({
  model: "Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo",
  messages: [
    { role: "system", content: "You are a senior software engineer. Write clean, production-ready code with error handling. Return only the code, no explanation." },
    { role: "user", content: "Write a TypeScript function that debounces an async callback, cancelling pending invocations when called again." },
  ],
  max_tokens: 1024,
});

console.log(response.choices[0].message.content);

Automated code review

Pass a diff and get structured feedback:

from openai import OpenAI
import json

client = OpenAI(
    base_url="https://api.runcrate.ai/v1",
    api_key="rc_live_YOUR_API_KEY",
)

diff = """\
-def process_payment(amount):
-    charge = stripe.Charge.create(amount=amount, currency='usd')
+def process_payment(amount, idempotency_key=None):
+    charge = stripe.Charge.create(amount=amount, currency='usd', idempotency_key=idempotency_key)
"""

response = client.chat.completions.create(
    model="deepseek-ai/DeepSeek-V3.2",
    messages=[
        {"role": "system", "content": "Review the code diff. Return a JSON array of objects with keys: line, severity (info/warning/error), comment."},
        {"role": "user", "content": f"Review this diff:\n\n```diff\n{diff}\n```"},
    ],
    max_tokens=1024,
)

reviews = json.loads(response.choices[0].message.content)
for review in reviews:
    print(f"[{review['severity'].upper()}] {review['comment']}")

Next steps

Extract structured data — use structured output with JSON schemas for reliable code analysis pipelines.
Route between models — use a fast model for linting and a strong model for complex refactors.
SDKs overview — Python, TypeScript, and Vercel AI SDK setup.