GMI Cloud

AI-native Inference Cloud.
Questions:

[GMI Cloud @gmi_cloud]

GMI 💚 @nvidia

[GMI Cloud @gmi_cloud]

Mark your calendar 🫵

📅 Wed 6/10, 4–5pm PT

we're going live on X and YouTube. @roangws will build and ship an AI agent from scratch using GMI AgentBox.

RSVP: https://t.co/TPcLwvJNDh

[GMI Cloud @gmi_cloud]

our team in denver: @venjia_z @lillian_ma_ @nicoleegong @YuhsinAChang @showmore234

[GMI Cloud @gmi_cloud]

We're in Denver this week for @CVPR!

Two events we're hosting:
→ Researcher demo night with @nvidia
→ Roundtable discussions on visual AI

If you are into agentic AI, AI infra, inference, or open-source, Come hang with the team at booth 434.

[GMI Cloud @gmi_cloud]

@TotallyNotParth @qwen_cloud soon

[GMI Cloud @gmi_cloud]

@CatGodSandHive @NVIDIAAIInfra

[GMI Cloud @gmi_cloud]

Just very proud to build AI Factory with @NVIDIAAIInfra

[NVIDIA AI Infrastructure @NVIDIAAIInfra]

At #NVIDIAGTC Taipei, our CEO Jensen Huang unveiled the AI infrastructure stack purpose-built for the agentic AI era:
→ NVIDIA Vera Rubin in full production
→ Vera CPU: the first CPU designed for agents
→ DSX platform: the playbook to create AI factories

Plus announcements across autonomous agents, physical AI, robotics, and more.

🔗 Read the recap: https://t.co/35p8Qd5Chf
📺 Watch the full keynote: https://t.co/u07hKWmmAH

[GMI Cloud @gmi_cloud]

@yuqih @lujomouton 🤣

[GMI Cloud @gmi_cloud]

@OpenRouter DS V4 flash here: https://t.co/X9ecwrhSpy

[GMI Cloud @gmi_cloud]

06.02.2026

we hit >110B daily tokens of DeepSeek V4 in @OpenRouter, >680B in the past week

Yipee!

[GMI Cloud @gmi_cloud]

05.26.2026

We hit >90B tokens of DeepSeek V4 in @OpenRouter, ~570B in the past one week!

Thank you everyone, this is just the start 🏋️

[GMI Cloud @gmi_cloud]

See you tmr at 7 pm 🗽

[GMI Cloud @gmi_cloud]

Hi New Yorker, we're co-hosting an AI agent sprint during #NYTechWeek

You'll ship a production-ready multi-agent app using @PhiniteAi 's Multi-Agent OS + GMI's inference, supported by @NVIDIAAI

Top 3 teams pitch to a VC. Food + drinks on us.

📍 132 Bowery, NYC
📅 Wed Jun 3, 7–11pm

RSVP here: https://t.co/gYcAAbfHIb

[GMI Cloud @gmi_cloud]

https://t.co/dKXy9ThroD

[GMI Cloud @gmi_cloud]

MiniMax M3 is now 25% off for 48 hours on GMI Cloud

input: $0.45
output: $1.80

[GMI Cloud @gmi_cloud]

@NVIDIAAIInfra proud to be a DSX platform building AI factories!

[GMI Cloud @gmi_cloud]

MiniMax M3 is now available on GMI Cloud

Go build: https://t.co/AeLuQv1RI1

[GMI Cloud @gmi_cloud]

Compared MiniMax M3 against three other frontier open-source models: DeepSeek V4, Qwen 3.7 Max, and MiMo 2.5 Pro

MiniMax M3 crashed others on:
1. collision physics
2. 3D pinball generation

It was the only model that produced a result for the 3D pinball test

For human motion rendering and hexagon sketching, Qwen 3.7 Max still tops

[GMI Cloud @gmi_cloud]

@JohnsonEnyenih2 @FireworksAI_HQ tyty

[GMI Cloud @gmi_cloud]

GMI Cloud 🤝 @FireworksAI_HQ

We're excited to partner with Fireworks AI to support scalable, production-ready open-model inference.

Their inference expertise + our high-performance NVIDIA infrastructure = an accelerated path from prototype to production.

Read more: https://t.co/QJe0pV7EdP

[GMI Cloud @gmi_cloud]

@BlockedPaths thank you~

[GMI Cloud @gmi_cloud]

Compared MiniMax M3 against three other frontier open-source models: DeepSeek V4, Qwen 3.7 Max, and MiMo 2.5 Pro

MiniMax M3 crashed others on:
1. collision physics
2. 3D pinball generation

It was the only model that produced a result for the 3D pinball test

For human motion rendering and hexagon sketching, Qwen 3.7 Max still tops

[GMI Cloud @gmi_cloud]

Honored to be an AI Cloud that is bringing AI factories closer to local industries and sovereign AI ecosystems of @nvidia

https://t.co/IQ4FXKBSZl