ATLAS Code

A coding model for generation, edits, refactors, and agentic tool use in your editor, on ElderAI’s OpenAI-compatible API.

Preview: open to every account with a verified email. Sign up, verify your email, and create an API key; new accounts get a free 7-day trial of 1,000 credits. No invite code needed. Behaviour, limits, and pricing may change before general availability.

Model id

atlas-code-1-preview

Base URL

https://elderai.cloud/v1

Endpoints

POST /v1/chat/completions (streaming and non-streaming), GET /v1/models; POST /v1/responses (Codex CLI, streaming and non-streaming)

Context

24,576 input + 8,192 output tokens per request

Tool calling

OpenAI tools / tool_calls

Price

65 credits / 1M input tokens · 16 / 1M cached input · 590 / 1M output (100 credits = $1)

Limits during the preview

• Invitation only. If GET /v1/models doesn't list atlas-code-1-preview, your account isn't in the preview yet.
• Cold starts. Capacity scales to zero when idle. The first request after a quiet period waits while a GPU worker starts: usually 1–2 minutes, and up to 3. After that, answers start streaming in about a second. Use streaming (stream: true; Codex always streams). While a request waits, the stream gets a : keep-alive comment every 15 seconds, which OpenAI SDKs and Codex ignore. Set client timeouts to at least 300 seconds.
• Shared capacity. Under load, requests may queue or return 429/503. Retry with backoff.
• Your plan's rate, concurrency, and context limits also apply. Thinking mode is off by default.

Your code is not used for training

We don't use your prompts, code, files, or completions sent to ATLAS Code to train or fine-tune any model. Requests are processed to return your answer and to meter usage. See the Privacy Policy.

Quickstart

Create an API key in your account and export it as ATLAS_API_KEY.

curl

curl https://elderai.cloud/v1/chat/completions \
  -H "Authorization: Bearer $ATLAS_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $(uuidgen)" \
  --max-time 300 \
  -d '{
    "model": "atlas-code-1-preview",
    "messages": [
      {"role": "system", "content": "You are a senior engineer. Answer with code."},
      {"role": "user", "content": "Write a Python function that parses ISO-8601 durations."}
    ],
    "max_tokens": 1024
  }'

OpenAI SDK — Python

import os
from openai import OpenAI

client = OpenAI(base_url="https://elderai.cloud/v1", api_key=os.environ["ATLAS_API_KEY"], timeout=300)
stream = client.chat.completions.create(
    model="atlas-code-1-preview",
    messages=[{"role": "user", "content": "Refactor this function to async: ..."}],
    max_tokens=2048,
    stream=True,
)
for chunk in stream:
    print(chunk.choices[0].delta.content or "", end="")

OpenAI SDK — TypeScript

import OpenAI from 'openai';

const client = new OpenAI({ baseURL: 'https://elderai.cloud/v1', apiKey: process.env.ATLAS_API_KEY, timeout: 300_000 });
const res = await client.chat.completions.create({
  model: 'atlas-code-1-preview',
  messages: [{ role: 'user', content: 'Write a Vitest test for a debounce() helper.' }],
  max_tokens: 1024,
});
console.log(res.choices[0].message.content);

Use it in your editor

Every tool needs the same three values: base URL https://elderai.cloud/v1, your ATLAS API key, and model atlas-code-1-preview.

Cursor

1. Go to Settings → Models → API Keys. Paste your ATLAS key into OpenAI API Key.
2. Turn on Override OpenAI Base URL and enter https://elderai.cloud/v1.
3. Click Add model, enter atlas-code-1-preview, enable it, and pick it in Chat.

Cursor sends custom-model requests from its own servers. Features that use only Cursor's own models, like Tab completion, won't use ATLAS Code.

Cline (VS Code)

1. Open Cline settings and set API Provider to OpenAI Compatible.
2. Set Base URL to https://elderai.cloud/v1, API Key to your ATLAS key, and Model ID to atlas-code-1-preview.
3. Set Context Window to 24576 and Max Output Tokens to 8192. Leave image support off.

Aider

export OPENAI_API_BASE=https://elderai.cloud/v1
export OPENAI_API_KEY=$ATLAS_API_KEY
aider --model openai/atlas-code-1-preview --timeout 300

OpenAI Codex CLI

Codex supports only the Responses API, and custom providers must set wire_api = "responses". ATLAS serves POST /v1/responses to every ATLAS Code account. Add this to ~/.codex/config.toml (Codex ignores provider settings in a project-level config):

model = "atlas-code-1-preview"
model_provider = "atlas"

[model_providers.atlas]
name = "ElderAI"
base_url = "https://elderai.cloud/v1"
env_key = "ATLAS_API_KEY"
wire_api = "responses"

The ATLAS Responses API is stateless (Codex's default store = false). It supports function tools and skips built-in tools such as web search.

Attribution

The ATLAS Code preview serves an open-weight base model, Qwen3-8B (Copyright 2024 Alibaba Cloud), licensed under the Apache License 2.0. ElderAI is fine-tuning its own version, which ships under the same name once it passes our tests; this notice will be updated then. Read the license text and notice. ATLAS Code isn't affiliated with or endorsed by Alibaba Cloud.