Hy3-preview API pricing on GMI Cloud
Every GMI Cloud Hy3-preview rate we track, compared against 3 providers serving the same model.
The cheapest Hy3-preview input price we currently track is $0.132/1M on OpenRouter.
GMI Cloud Hy3-preview rates
Prices per 1M tokens. Batch and cached-input rates are shown where the provider publishes them.
| Mode | Input | Output |
|---|---|---|
| Standard | $0.140/1M | $0.580/1M |
About Hy3-preview
- Creator
- Tencent
- Modalities
- text
- Open weights
- Yes
Hy3-preview on other providers
About GMI Cloud
GMI Cloud is a neocloud combining dedicated NVIDIA GPU compute (H100, H200, B200, GB200, with GB300 on pre-order) and a serverless, OpenAI-compatible inference API for text, image, video, and audio models.
Frequently Asked Questions
How much does Hy3-preview cost on GMI Cloud?
GMI Cloud serves Hy3-preview from $0.140 per 1M input tokens, across 1 tracked pricing mode. Prices are collected daily — see the table above for current input, output, and cached-input rates.
Is GMI Cloud the cheapest way to run Hy3-preview?
GMI Cloud ranks #2 of 3 providers we track serving Hy3-preview. The cheapest input price right now is on OpenRouter. Price is only one factor — throughput, latency, and context limits differ between providers.
Does GMI Cloud offer batch or cached pricing for Hy3-preview?
We track 1 Hy3-preview offering from GMI Cloud. Batch and cached-input rates appear in the table above when the provider publishes them; blank cells mean we have no current figure for that mode.