one cloud for compute, inference, and agents Question: https://xitter.cf/t.co/dOcscX4hmx Japan: @gmicloudJapan
Mountain View, CA
Joined December 2023
- Tweets3.6K
- Following62
- Followers10.2K
- Likes3.4K
Pinned Tweet
Announcing our $668M Series B!
Led by ARCHIV with participation from @nvidia.
This capital expands our GPU capacity across the U.S., Taiwan, and APAC, and scales our inference platform.
Our contracted ARR has reached more than 9x since the end of 2025.
Thanks to everyone who is building with us 🙏
our post in May: xitter.cf/gmi_cloud/status/20502…
We are proud to support the @NVIDIAAI OpenShell Reference Architecture and bring it to our infrastructure.
Building on GMI Cloud’s investment in the NVIDIA Vera Rubin platform: model compute, governed agent execution and hardware-level security are all on one cloud.
Agents can work with real data, tools and services, with clear permissions, runtime controls and a full record of what they do.
🤖 Made with AI
AI agents need fast tool calls and security controls they can’t rewrite.
NVIDIA OpenShell Reference Architecture combines the compute to run agents at scale with runtime controls that govern their access to tools and data.
See how the hardware and software work together: nvda.ws/4BiJBfY
read more: gmicloud.ai/en/blog/gmi-clou…
The 5 second video from that session.
Prompt: a paper boat drifting on a calm pond, soft morning light.
Model: seedance-1-0-pro-fast on GMI Cloud, 1248 x 704, ready in about 27 seconds. gmi saved it to ./gmi-output/ and printed the path.
Day 2 of gmi: media from inside the session.
/image and /video, with the price shown before anything runs.
Recorded live. The waits are shortened but the numbers are real: a $0.035 image, then a 5 second video at $0.022 per second.
from @zkasuran
We put EAGLE-4 on NVIDIA B300 and benchmarked it against every other speculator for Qwen3.8-27B
@EagleCorp
1. 703 tok/s at batch
2. 2.7× faster than plain decoding
3. 27,600 tok/s on one 4×B300 replica at 512 concurrent
Full breakdown here 👇
Proud to be a launch partner for EAGLE-4, live on GMI Cloud from day one.
We’ve been running it on our NVIDIA B300 fleet: 2.7× faster than plain decoding on Qwen3.8-27B.
Live on GMI Cloud today. Full benchmarks coming soon!
EAGLE-4 is here.
From EAGLE-1 to EAGLE-3, we’ve kept making inference faster.
EAGLE-4 is our fastest yet.
961.6 tokens/s on @Alibaba_Qwen’s Qwen3.8-27B with TokenSpeed (BF16, 4× @NVIDIA B300, TP4, one request, T=0).
Github: github.com/eagle-hq/EAGLE-4
GMI Cloud retweeted
GMI Cloud gave us free credits to give Qwen 3.8 Max an honest review on our tool calling bench. The Tīrtha Bench review: it's very good. With reasoning off it scored 96.1% on 800 tasks and passed 100 of 100 trick requests, level with Opus 5.5 at about a quarter of the cost per right answer. Under load, 99% of tool calls came back within 4 seconds, with zero errors in 1,840 calls. So we're moving Tīrtha's top level to it. Thanks @gmi_cloud and @Alibaba_Qwen. tirtha.ai/bench
Day 1 works today.
Type gmi and you're in a session:
/ opens every command
/model switches models mid-conversation
/models shows the Model Hub with prices
Answers stream in as formatted text with code blocks
Scripts get the same tool: gmi chat "..." --json
Demo at 2 PM PT.
The code is written by a coding agent running GPT-6.1 Sol on GMI Cloud, the same API the CLI talks to.
Code and install: github.com/zkasuran/gmi
Tomorrow: /image, /video and /audio, without leaving the session.
