engyThe fuel for a world run by agents.
Every agent, every step, runs on inference. engy turns frontier open models into that fuel: abundant, low-cost, and built to scale with the autonomous economy taking shape.
- Reliable and agentic-native. Served from regulated datacenters, 7×24 on-call, tuned for agentic traffic. High cache-hit rates and low cost at scale.
- Zero data retention. Your prompts and outputs are never stored or trained on, on every model marked ZDR below; terms has the detail.
- SOTA open models. Full and original weights. Frontier models like GLM-5.2 and Kimi K3, with new models added as they ship.
Models
See full pricing →| model | input | output | cached | ctx |
|---|---|---|---|---|
| qwen3.8-27bnewZDR | $0.045 | $0.32 | $0.015 | 1M |
| deepseek-v4-flash-0731newZDR | $0.045 | $0.09 | $0.009 | 1M |
| glm-5.2ZDR | $0.68 | $1.50 | $0.18 | 262K |
| kimi-k3ZDR | $1.50 | $7.50 | $0.15 | 1M |
| qwen3.6-35b-a3b | $0.045 | $0.30 | $0.015 | 208K |
| $ per 1M tokens | ||||