CTO @conste11ation | generating utility

Joined November 2015
21 Photos and videos
Pinned Tweet
AI has crossed a threshold and knowledge work will never be the same. Last week we announced Gate AI, drop-in security and token savings for any agent. Here's why we built it: constellationnetwork.io/blog…
20
41
107
67,613
"Trust me, I checked the logs." That's the best answer most of us can give about what an OpenClaw agent did on a long unattended run. And the log is a plain text file on disk that anything can edit, including the agent. So we built Gate OC Audit: a plugin that records every tool call, message, skill, and cron to a tamper-evident trail on your own machine. One line to install, and it hooks into the harness directly. Query it from the CLI, browse it in a local dashboard, set alerts for first-seen tools and unusual spikes. Everything stays on your disk. No account, no cloud. If you ever need to prove the record to someone else, anchor it to Digital Evidence for free and anyone can verify it without trusting you. Or us. Full story: constellationnetwork.io/blog… Plugin: constellationgate.ai/opencla…
10
29
152
1,082,915
Fable is a very capable model and a step up from Opus - I’ve been testing it all afternoon. The larger implication though is that this is the pivot point tipping towards the end of the monthly subscription model. SOTA models will be sold at their business value, tens of thousands a month or more. Open source models will rapidly catch up to Opus and be available at a fraction of the cost. Local models will rapidly advance also to fulfill most day-to-day needs. How much intelligence do you need? And what’s your budget?
The release of Claude Fable-5 just accelerated the intelligence class divide. Upgraded intelligence, expensive APIs, heavier censorship, and subscription limits that drain way faster. The 5-hour limit on the $200 plan is gone in less than an hour. Moving forward, the productivity gap between devs who use this and those who can't afford it will be dozens of times wider. This applies to companies, too. Microsoft previously banned Opus due to the cost, but now this gap will force everyone to use much more expensive frontier models. Welcome to the permanent underclass.
2
30
1,070
Alex Brandes ² retweeted
Replying to @Conste11ation
@Conste11ation launches @arca_wallet, a dollar wallet anyone can open in about 30 seconds. Save, send, and spend digital dollars from your phone, with yield powered by @labs_compass.
2
9
31
2,586
Slowly, then all at once
Slowly, then all at once. $DAG
4
10
52
2,750
Frontier models become marginally better at massive cost over the next year. Opus-level intelligence shifts to low cost models at the same time. Everyone runs overall more compute but optimized to the cost/intelligence necessary for a task. Forget AGI - think hyper-compute.
Good take My guess is - demand for intelligence is near infinite - but 80% of workloads will be running on 99% cheaper models within 12-18 months - 20% of workloads will still run on latest gen models where IQ maxing is important (scientific breakthroughs, higher level ochestrator agents?) - rough analogy might be what % of macbooks or gaming PCs sold have the maxed out specs for CPU/GPU, prices are falling much faster than Moore's law here though - this leads me to think the limiting factor will be energy and compute, not better models At Coinbase we're working hard on routing prompts to cheaper models where appropriate, and in some cases have been able to keep costs roughly flat, while token usage continues to grow exponentially.
5
32
1,192
You could also frame this as American AI startups are routing far more traffic to open source models than closed source ones. Historically, open source has always won.
American AI startups are routing far more app traffic to Chinese LLMs. Through OpenRouter, weekly token consumption was mostly driven by U.S. models through much of 2025. But from early 2026, Chinese models suddenly became the main growth engine. AI model market is becoming less about brand loyalty and more about raw utility.
3
16
588
If you’re shocked by Apple’s announcement today that Apple Intelligence will run licensed models from Google Gemini, you’re missing the bigger picture. Apple is investing in building the most advanced local AI architecture available - they’re competing with Nvidia, not Google.
Apple's entire brand is building everything itself. Today it handed the keys to Google. The next generation of Apple Intelligence now runs on a custom 1.2 trillion parameter Gemini model. The most vertically integrated company on earth just outsourced its smartest layer. #WWDC
1
1
22
1,272
While everyone focuses on frontier model development, the local model race keeps making major gains. Models of this size can easily fit on most consumer hardware with reasonable performance and zero subscription cost. We’re seeing more useful AI that people can actually own, run, and build around every day.
Meet Gemma 4 12B! A unified, encoder-free multimodal model designed to bring high-performance intelligence directly to your laptop, and released under an Apache 2.0 license. Bridging the gap between edge efficiency and advanced reasoning. Here is what’s new with Gemma 4 12B: 👇
1
5
36
1,271
Alex Brandes ² retweeted
Replying to @arca_wallet
@arca_wallet (by @Conste11ation) is live. Send, save, and spend dollars from your phone, then verify once and tap a growing network of partner apps from the same wallet. @labs_compass powers the DeFi layer. Congrats to the Arca team!
Arca is here. Send, save, and spend real dollars from your phone. No bank account, no paperwork, no crypto knowledge required. Live on iPhone and Android.
2
7
27
3,122
Arca wallet is officially live!
We built Arca Wallet as a self-custodial dollar wallet. But the marketplace turns it into something bigger. Verify once, unlock a growing network of partner apps from the same place you hold and send money. Live now on iOS and Android. finance.yahoo.com/markets/cr…
4
12
89
73,352
Today the team and I finally get to reveal what we've been quietly building: Constellation Gate AI, the prompt-injection defense the developer market has been missing. It works, and we have the benchmarks to prove it. The state of AI agent security right now: prompt injection is OWASP's #1 LLM threat, attacks against AI coding agents like Claude Code and OpenClaw are increasing in the wild, and the only commercial defenses are sold as six-figure annual contracts to enterprise. If you're an individual developer or a small team running an AI coding agent, there is no AI security solution within reach. We built Gate to fill that gap - a drop-in security layer between your AI agent and the model. It screens every request and response for prompt injection, credential leaks, and hijacked tool calls. Available self-serve and bring your own key, so it works with your existing Claude or chatGPT subscription. The proof: we published the benchmark methodology and results today. 97.4% F1 across 16 public prompt-injection datasets. #1 on 8 of them. On the matched comparison with Lakera Guard, the leading enterprise tool, we land at 96.6% F1 to their 83.7%, with false-positive rates 6 to 17 times lower on legitimate prompts. Read the blog post: constellationnetwork.io/blog… Sign up for early access to Gate: constellationgate.ai/ @Conste11ation #AI #OpenClaw
AIAI Holdings' Constellation Network Unveils Gate AI Security Gateway and Performance Benchmarks Ahead of June Launch We benchmarked Gate AI across 16 public prompt-injection datasets. Ranked #1 on half. 53ms median latency. 6-17x fewer false positives than Lakera Guard on benign traffic. One global threshold, no per-dataset tuning. accessnewswire.com/newsroom/…
25
64
170
13,395
Wow - the Alibaba team continues to deliver at rapid pace. Qwen3.6 is already the most balanced cost-to-intelligence model series available to open source. It also powers most of the consumer AI features offered by Silicon Valley today 👀 I’m excited to give 3.7 Max a try.
📣Meet Qwen3.7-Max — our latest flagship, made for the Agent Era. A versatile foundation for agents that actually get things done: 🧑‍💻 Coding agent, end to end. Frontend prototypes, multi-file refactors, real debugging — nails it. 🗂️ A reliable office and productivity assistant. Get your work done through MCP integrations and multi-agent orchestration. ⏱️ Long-horizon autonomy. 35 hours straight on a kernel optimization task — 1,000 tool calls, zero hand-holding. 🔌 Scaffold-agnostic. Claude Code, OpenClaw, Qwen Code, or your own stack. Consistent reliability everywhere. API's up on Alibaba Model Studio. You can also take it for a spin on Qwen Studio. Go build something wild!🏃🏃‍♂️ 📖 Blog: qwen.ai/blog?id=qwen3.7 ✅ Qwen Studio: chat.qwen.ai/?models=qwen3.7… ⚡️ API:modelstudio.console.alibabac…
1
25
1,285
Alex Brandes ² retweeted
AIAI Holdings Corporation highlights Constellation Network. Constellation Network is a technology company that includes a Layer 1 blockchain, enabling AIAI to expand its portfolio into blockchain-enabled infrastructure. Read more here: buff.ly/4ZSLdXv
12
46
116
13,074
Alex Brandes ² retweeted
⚠️Phishing alert. A fake email is circulating from with a "Claim $DAG" button. This is not from us. There is no $DAG rewards program. There is no community airdrop. Constellation Network will never ask you to connect your wallet to claim tokens. Do not click. If you received this email, delete it.
5
29
120
5,233
Alex Brandes ² retweeted
Opened an account in seconds and received my usdc transfer instantly. Nice work 👏🏻 @arca_wallet / @Conste11ation
4
4
41
1,103
While I love Claude Code, I think the writing is on the wall that the subsidies won’t last forever. They continue to exclude “undesirable” usage patterns from their service. Innovation will thrive elsewhere. Luckily, there are a ton of great providers out there! Shout out to Codex for not playing these games but open source and local LLMs are really the ones to watch 👀
Starting June 15, paid Claude plans can claim a dedicated monthly credit for programmatic usage. The credit covers usage of: - Claude Agent SDK - claude -p - Claude Code GitHub Actions - Third-party apps built on the Agent SDK
1
23
1,106
Alex Brandes ² retweeted
Fastest dollar account creation in the world!
2
22
45
5,415
True adoption happens when the technology fades into the background. @arca_wallet is a stablecoin wallet for the masses that requires no prior crypto knowledge to use.
Most crypto apps feel like a technical chore. That’s because builders are forced to juggle infrastructure instead of perfecting their product. This is why we build. We’re proud to see @arca_wallet using ZeroDev and @Privy to turn complex onchain hurdles into a flow that feels inescapably normal. Stop managing gas and chains. Start building what you couldn’t before.
12
61
1,471
Alex Brandes ² retweeted
May 12
We started closed testing for @arca_wallet yesterday. Honestly the biggest thing so far has been seeing people who normally get scared away by crypto wallets actually use it without needing help. Still early, but feedback has been great so far. Shoutout to @privy_io, @labs_compass and @zerodev_app for the amazing service that is enabling this smooth experience!
4
18
89
5,358