Sabitlenmiş Tweet
Honored to support and work with @BakerjBrandon to deliver a world class solution against big tech and their centralized infra. @conduit_network changing paradigms!
English
Scott
191 posts
Scott
@scottychain
Co-founder of @conduit_network / Ex-Consensys I talk about DePIN, AI, Dev Tooling & Infra that Scales
Katılım Kasım 2021
100 Takip Edilen24.6K Takipçiler
Big names at this event!
Brokk, Inc.@BuildWithBrokk
Tune in this Thursday at 3.00pm PST for our X space on the Rise of Agentic AI. Hosted by @Dylen_Turnbull, we have the pleasure of being joined by @mikegchambers from @awscloud, @JMHReif from @neo4j and @Cervator from @nvidia along with @spyced
English
@UnslothAI Let's hope that Kimi gets bumped up in the Power Ranking - 2.0 didn't do so hot. brokk.ai/power-ranking
English
You can now run Kimi K2.5 locally! 🔥
We shrank the 1T model to 240GB (-60%) via Dynamic 1-bit.
Run at >40 tok/s on 240GB VRAM/RAM.
2-bit is recommended as it passes our code tests. Run near full precision on 622GB.
Guide: unsloth.ai/docs/models/ki…
GGUF: huggingface.co/unsloth/Kimi-K…
Kimi.ai@Kimi_Moonshot
🥝 Meet Kimi K2.5, Open-Source Visual Agentic Intelligence. 🔹 Global SOTA on Agentic Benchmarks: HLE full set (50.2%), BrowseComp (74.9%) 🔹 Open-source SOTA on Vision and Coding: MMMU Pro (78.5%), VideoMMMU (86.6%), SWE-bench Verified (76.8%) 🔹 Code with Taste: turn chats, images & videos into aesthetic websites with expressive motion. 🔹 Agent Swarm (Beta): self-directed agents working in parallel, at scale. Up to 100 sub-agents, 1,500 tool calls, 4.5× faster compared with single-agent setup. - 🥝 K2.5 is now live on kimi.com in chat mode and agent mode. 🥝 K2.5 Agent Swarm in beta for high-tier users. 🥝 For production-grade coding, you can pair K2.5 with Kimi Code: kimi.com/code - 🔗 API: platform.moonshot.ai 🔗 Tech blog: kimi.com/blogs/kimi-k2-… 🔗 Weights & code: huggingface.co/moonshotai/Kim…
English
This will be a great discussion!
Brokk, Inc.@BuildWithBrokk
Everyone says LLMs are the future, but few are talking about the cost of running them in 2026. Join us live on X spaces in 24 hours with guests Jonathan Ellis @spyced (Brokk/DataStax), Stephen Blum @stephenlb (PubNub), and Paul Pindell @ppindell (F5). Hosted by @DylenTurnbull
English
I can't stop building games with my 7-year-old using Claude Code.
Our latest is a retro pixel space shooter:
→ 3 worlds (space, desert, hell)
→ Boss battles at the end of each level
→ Power-ups and wave-based enemies
We're living in a world where anyone can rebuild the games they loved as a kid with AI.
📌 Play it here: space-pixel-shooter.vercel.app
English
@0xmitsurii Check out Brokk.ai - former Cassandra Founder created a way to simply supervise agents to code
English
@EHuanglu Watch this AI Dev tool fix a real precision bug in Blender's 21k codebase - youtube.com/watch?v=t2E6AO…
YouTube
English
So much easier to see which model works best for your situation!
Brokk, Inc.@BuildWithBrokk
Stop playing LLM roulette. Brokk Power Rankings tells you exactly which model wins on your tasks.
English
BREAKING: Grok Code Fast 1 wins the #1 spot on this week’s OpenRouter AI Model Rankings Leaderboard with ~420 billion tokens usage. 🔥
English
@aidenybai Check out Brokk - those guys have figured out large code bases
English
alright, fuck it
i'm going to try every single code review bot to see which one's the best
bugbot, greptile, coderabbit, graphite...?, who else am i missing
Aiden Bai@aidenybai
why does cursor's bugbot not find all issues on the first try? i keep having to bugbot run → fix in cursor → bugbot run → fix in cursor → bugbot run over and over again
English
@MichaelHiles Many L1's shouldn’t claim custody because of their security flaws
English
You know that feeling when you ask an AI to refactor your auth system and it comes back with 47 changed files and you have no idea what just happened?
Yeah. We felt that too.
That's why our top engineer Lutz started doing something different.
First, he'd make the AI research - "show me what you found about how auth works in this codebase"
Then, plan - "ok, break this into actual achievable tasks"
Only then would he let it build - and he'd redirect it the moment it went sideways.
It worked so well we built it into Brokk. Now everyone can work like Lutz.
Research → Plan → Build. You're in the driver's seat at every turn.
English
@BuildWithBrokk This workflow changes everything! Lutzmode for the W!
English
Paying more doesn’t always mean you’re getting more…
Brokk, Inc.@BuildWithBrokk
You're probably overpaying for AI coding. We tested every major model. If you're still paying full price for Claude or GPT-5, you're leaving money on the table. If cost matters, there are only two models worth your attention. And one of them will surprise you. 🧵
English
