GMICloud

GMICloud Pricing

Visit gmicloud.ai →

Per-token pricing for 0 open-source LLMs on GMICloud — a GPU cloud and serverless inference provider serving frontier open models like DeepSeek V4 and GLM 5.

Last updated: Nov 21, 2026

Get our weekly newsletter on pricing changes, new releases, and tools.

GMICloud offers free models

$5 credit + free models — see the full GMICloud free tier guide

No GMICloud pricing data available.

About GMICloud

GMICloud is a GPU cloud and inference platform offering serverless, pay-per-token access to large open-source models. Its catalog leans into frontier open releases — DeepSeek V4, GLM 5, Qwen3.5 and Tencent Hunyuan — often served at fp8 or bf16 with very long context windows (up to 1M tokens). The API is OpenAI-compatible.

GMICloud also rents dedicated GPU capacity for teams running their own deployments. The table below shows the models it currently serves with public per-token pricing, the quantization each price applies to, and context length.

GMICloud Pricing FAQ

How much does GMICloud cost?

GMICloud charges per token, with input prices for large open models starting around $0.06–$0.30 per million tokens. See the live table above for current per-model input and output prices.

Which models does GMICloud support?

Frontier open-source LLMs including DeepSeek V4, GLM 5, Qwen3.5 and Tencent Hunyuan. The table above lists the models GMICloud currently serves with public per-token pricing.

Does GMICloud support long context?

Yes — several GMICloud models offer very long context windows, up to roughly 1M tokens. The Context column in the table shows the limit for each model.

Is GMICloud OpenAI-compatible?

Yes. GMICloud exposes an OpenAI-compatible API, so switching from another provider is usually just a base-URL and API-key change.