Derivatives exchange group for compute, inference, and AI commodities. Offering the American Innovation Exchange, Liquid Inference, and AX.
Chicago, IL
- Tweets 578
- Following 194
- Followers 14,422
- Likes 1,201
Pinned Tweet
Today we’re launching Liquid Inference, an OpenAI/Anthropic-compatible router where inference providers compete on price for every prompt. Buyers route across 700+ open- and closed-weight models at the lowest available marginal cost, providers onboard instantly to start serving models.
8
22
135
33,777
Architect retweeted
Liquid CLI + Oh My Pi is my go-to local harness setup now. It takes 3 commands to have everything ready under 5 minutes
- Install Liquid CLI
- liquid login
- liquid omp
Liquid Inference has copy-paste setup for 21 tools, including Claude Code, Codex, Cline, OpenCode, LiteLLM, LangChain and the Vercel AI SDK.
For Claude Code it's two environment variables.
2
3
16
721
Architect retweeted
I’m honored to speak at the @CFTC’s first Frontier Forum in DC next week. Looking forward to discussing the role of SI in US futures markets. Thank you to Chairman Selig and the Innovation Task Force.
Super intelligence is poised to transform our financial markets.
Through the @CFTC's inaugural Frontier Forum, we are bringing together leaders at the intersection of technology and finance to explore the opportunities and risks of super intelligence and help shape the next frontier of American financial innovation.
Read more about our first Frontier Forum on October 28:
cftc.gov/PressRoom/PressRele…
8
9
75
3,294
Liquid Inference has copy-paste setup for 21 tools, including Claude Code, Codex, Cline, OpenCode, LiteLLM, LangChain and the Vercel AI SDK.
For Claude Code it's two environment variables.
5
6
32
2,751
GPU COMPUTE WEEKLY SNAPSHOT
10-19-26
$/GPU-hr · 7D · 1M · YTD · 30D ann. vol
>H100 — $2.91 · +1.60% · +7.29% · +55.5% · 9.6%
>H200 — $3.54 · +2.78% · +3.57% · +45.8% · 11.0%
>B200 — $5.74 · +5.67% · +11.82% · +48.2% · 8.0%
>B300 — $7.60 · +2.45% · +27.32% · +55.5% · 12.2%
1
4
947
Architect retweeted
More Liquid Inference integrations launched over the weekend:
- @Deepseek_ai V4.1-Flash-1017 model support and Deepseek Harness integration
- @Openclaw plugin for agent backend
- @ComfyUI plugin for image and video generation and editing
- Embedding models
We ship every day.
8
5
42
3,188
Architect retweeted
What if AI inference were priced like stocks?
@Architect_Fi Founder and CEO @BrettHarrison explains how an auction-based marketplace could drive down the cost of running AI models:
“In our version of this sort of competitive inference exchange, people can constantly move their prices around based on their marginal cost.”
11
13
55
14,615
Architect retweeted
With open-weight model use surpassing frontier models among developers, even ahead of the IPOs, inference tokens are now behaving like commodities.
What if AI inference were priced like stocks?
@Architect_Fi Founder and CEO @BrettHarrison explains how an auction-based marketplace could drive down the cost of running AI models:
“In our version of this sort of competitive inference exchange, people can constantly move their prices around based on their marginal cost.”
13
10
88
5,552
Architect retweeted
Alpha.sh is live on Liquid Inference.
The Liquid Inference router network is growing quickly. We now have 28 verified companies serving >1000 models, offering capacity dynamically and updating quotes in real-time. Average time for providers to onboard and integrate models is 3 hours, full verification within a day.
2
1
9
369
Let there be tokens 🚀
The Liquid Inference router network is growing quickly. We now have 28 verified companies serving >1000 models, offering capacity dynamically and updating quotes in real-time. Average time for providers to onboard and integrate models is 3 hours, full verification within a day.
3
11
378
Architect retweeted
What's the smartest model you can get for $0.05 per million tokens?
We built a page that answers that. Today it's GLM-5.3: 2 points behind Claude Sonnet 5.5 on the Artificial Analysis index, for about 1/100th the price.
Prices are live from the exchange. Try your own budget: liquidinference.ai
16
9
63
2,631
Architect retweeted
Excited to see @jetinfer joining @Architect_Fi's inference provider network!
We're focused on delivering fast, cost-efficient inference for open-weight models, and it's great to collaborate with teams pushing this ecosystem forward.
Thanks to Brett and the Architect team for having us on board! 🚀
The Liquid Inference router network is growing quickly. We now have 28 verified companies serving >1000 models, offering capacity dynamically and updating quotes in real-time. Average time for providers to onboard and integrate models is 3 hours, full verification within a day.
7
12
914
Architect retweeted
Liquid Inference now supports images and video! Video inference costs aren’t optimized to the same level as text, and a fair auction brings iterative AI content creation in line. Access via our API or check out the compare tool in our web playground.
8
8
59
3,927
Architect retweeted
Including us! U want tokens? I got em!
Great to work with liquid - let’s bring everyone more token abundance @BrettHarrison
The Liquid Inference router network is growing quickly. We now have 28 verified companies serving >1000 models, offering capacity dynamically and updating quotes in real-time. Average time for providers to onboard and integrate models is 3 hours, full verification within a day.
1
8
718
Architect retweeted
And you thought openrouter was the ultimate solution.
Here inference providers compete for your queries.
Compute markets.
The Liquid Inference router network is growing quickly. We now have 28 verified companies serving >1000 models, offering capacity dynamically and updating quotes in real-time. Average time for providers to onboard and integrate models is 3 hours, full verification within a day.
2
3
31
2,108
Architect retweeted
Liquid & Permissionless Commodities empower everyone to eat every software.
Our role is to ensure access to censorship resistant and accelerated intelligence.
The future is liquid, the future is instant.
The Liquid Inference router network is growing quickly. We now have 28 verified companies serving >1000 models, offering capacity dynamically and updating quotes in real-time. Average time for providers to onboard and integrate models is 3 hours, full verification within a day.
6
19
1,039
Architect retweeted
Excited to be one of the new verified providers on the Liquid Inference network by @Architect_Fi
Inference is rapidly becoming commoditized and financialized, and Liquid Inference is well positioned to bring real-time price discovery and continuous competition to every token.
The Liquid Inference router network is growing quickly. We now have 28 verified companies serving >1000 models, offering capacity dynamically and updating quotes in real-time. Average time for providers to onboard and integrate models is 3 hours, full verification within a day.
1
8
28
1,924
Architect retweeted
The Liquid Inference router network is growing quickly. We now have 28 verified companies serving >1000 models, offering capacity dynamically and updating quotes in real-time. Average time for providers to onboard and integrate models is 3 hours, full verification within a day.
16
28
189
22,339
Welcome to the Liquid Inference network, @thinkinstant!
It was a pleasure working with @BrettHarrison and his team on our first deployment of Instant Inference.
We are grateful for @Architect_Fi's dedication to serving quality liquid commodities such as Instant.
Accelerate, no one can stop you now!
1
8
34
3,195