Models

Pricing for every ATLAS model. Generate a scoped API key for any model.

⚡

HelixFree

deepseek/deepseek-v4-flash

Fast and capable. The free tier model — great for everyday tasks, quick answers, and high-volume use.

Context window1M tokens

Input cost$0.30 / 1M tokens

Output cost$0.60 / 1M tokens

Available onFree, Pro, Master

🌟

VegaPro

qwen/qwen3-coder-30b-a3b-instruct

A powerful mid-tier model tuned for coding, analysis, and complex reasoning. The workhorse for serious builders.

Context window128K tokens

Input cost$0.45 / 1M tokens

Output cost$3.00 / 1M tokens

Available onPro, Master

🐍

ViperMaster

deepseek/deepseek-v4-pro

Our most powerful model. Unrestricted reasoning, massive context, and the highest quality output available.

Context window1M tokens

Input cost$1.30 / 1M tokens

Output cost$2.60 / 1M tokens

Available onMaster only

👁

ATLAS VisionPro

atlas/vision-v1

Multimodal image + text understanding. Analyze screenshots, diagrams, photos, and documents.

Context window128K tokens

Input cost$0.50 / 1M tokens

Output cost$1.00 / 1M tokens

Available onPro, Master

🎤

Voice ModeMaster

atlas/voice-v1

Real-time duplex voice conversation. Ultra-low latency speech synthesis and understanding.

Input cost$0.01 / minute

Output cost$0.02 / minute

Latency<300ms

Available onMaster only

🔊

Text to SpeechMaster

atlas/tts-v1

Natural voice synthesis across 32 voices and accents. Output in MP3, WAV, or OGG.

Cost$0.015 / 1K chars

Voices32 available

Output formatsMP3, WAV, OGG

Available onMaster only