Scalable Compute.
Autonomous Intelligence.
Choose the compute allowance and agent swarm capacity tailored for your engineering workflow. From personal rapid coding to multi-agent production swarms.
Aether Free
Standard exploration and basic developer intelligence assistance.
- 50 mX Credits monthly
- Standard model routing
- Web chat interface
- Basic Terminal pairing
- No overage rate discounts
Aether Pro
High-speed GPU execution for full-stack developers and builders.
- 1,000 mX Credits per month
- 50% Overage Discount ($0.003/1k)
- Autonomous CLI file manipulation
- High-speed GPU inference stream
- Full-context code generation
Aether Ultra
Peak reasoning capacity for large codebases and complex architectures.
- 1 Month Free Included — $0 first 30 days, full 5,000 mX quota immediately
- 5,000 mX Credits per month
- 75% Overage Discount ($0.0015/1k)
- Maximum Priority Queueing
- Deep technical plan & audit modes
- Unlimited CLI pairing tokens
Aether Business
Autonomous multi-agent swarms, shared secret vaults, and governance.
- 30,000 mX Credits + Custom pool
- Aether Team Swarms (Product, QA, Ops)
- Secure Team API Key Vault
- Zero-Data Retention SLA Guarantees
- Dedicated GPU routing & Support
Compare Plan Capabilities
Detailed breakdown of quotas, compute allocations, and execution environments.
| Feature & Metric | Free | Pro | Ultra | Business |
|---|---|---|---|---|
| Monthly mX Compute Credits | 50 mX | 1,000 mX | 5,000 mX | 30,000+ mX |
| Overage Cost per 1,000 Tokens | $0.0060 | $0.0030 (50% Off) | $0.0015 (75% Off) | Custom Contract |
| Aether Code Autonomous Terminal | Standard | Full Native Access | Full Native Access | Multi-Agent Swarm |
| GPU Inference Priority | Standard Tier | Accelerated | Top-Tier Priority | Dedicated Queue |
| Autonomous Multi-Agent Team (Aether Team) | — | — | Single Leader | Full Swarm Active |
| Encrypted Team Secrets Vault | — | — | — | Included |
| Privacy & Data Retention | Opt-out Available | Opt-out Available | Zero-Data Option | Strict Zero-Data SLA |
Sub-100ms TTFT
Powered by high-throughput tensor parallel pipelines, delivering ultra-low latency token streams directly into your terminal.
Zero-Data Retention
Enterprise privacy by design. Your code, filesystem changes, and workspace context are never stored or used for model training without consent.
Autonomous CLI Pairing
One-click 6-digit PIN connect pairs your local IDE and command line with full read, write, and command execution capabilities.