Ilusha Help

Creator · 9.4M subscribers
Promoting brands · millions of views
Building apps & systems

[Ilusha Help @ilushahelp]

⚡🔥 QWEN3.5-9B JUST OUT-CAUGHT THE FRONTIER ON FRAUD 🔥⚡

🚨 COINBASE FINE-TUNED A SMALL OPEN-WEIGHT MODEL ON ITS OWN ONRAMP FRAUD DATA
📊 VS OPUS 4.5: +9.6 PTS F1 • +12.0 PTS RECALL • +35.4 PTS DOLLAR-WEIGHTED RECALL
🧪 16,140-TRANSACTION BENCHMARK
⏱️ MEDIAN LATENCY 0.683s VS 1.515s (~55% FASTER)
🧠 BRIAN ARMSTRONG: SPECIALIST MODELS WILL BEAT GENERALISTS ON NARROW TASKS

The new playbook: frontier models for breadth, your own fine-tuned small model for the job that pays.

🔗 https://t.co/5L4krpObIq

#AI #Qwen #Coinbase #OpenSource #FineTuning #LLM

[Brian Armstrong @brian_armstrong]

We were able to fine tune an open weight model (Qwen3.5-9B) on proprietary internal data at @coinbase, and it out performed frontier models, at much lower cost and speed, for a specialized task (fraud prevention in this case).

I think we'll see more of this pattern inside companies. The frontier models are amazing generalists. But increasingly you'll see specialist models out perform on more narrow tasks, either working on teams, or by themselves.

We're making it easy to fine tune models internally for our employees. We even have a skill for it in our internal harness that makes it easy to spin these up and start training them:

"/fine-tune-and-serve-llm I want to fine-tune Qwen for payments risk review. help me get started"

I imagine we'll have some backend services that could be replaced entirely by this new architecture.

[Ilusha Help @ilushahelp]

⚡🔥 CARBON-A — DNA IN → GENE MAP OUT 🔥⚡

🚨 HUGGING FACE BIO JUST DROPPED AN OPEN GENE FINDER
🧬 1.2B MODEL READS RAW DNA → PROTEIN-CODING REGIONS
🌍 566M CANDIDATES ACROSS 22,617 SPECIES (MANY NEVER STUDIED)
📊 ONE MODEL FOR MAMMALS → PLANTS → FUNGI • nt F1 0.944
🧪 WET-LAB CHECKS (Iso-Seq) BACK SOME “NEW” GENES VS RefSeq
📦 OPEN WEIGHTS • Apache 2.0 • CADB Explorer live

Top AI × bio chatter: sequencing got cheap — annotation is the bottleneck. Primary is HuggingFaceBio.

🔗 https://t.co/rs0lwH6qHM
🔗 https://t.co/bMu2Jq19vk

#AI #BioAI #CarbonA #HuggingFace #OpenSource #Genomics

[Ilusha Help @ilushahelp]

⚡🔥 ANATOMY — ONE LINE → A LIVE MACHINE YOU CAN EXPLORE 🔥⚡

🚨 NEW OPEN-SOURCE CLAUDE CODE SKILL IS TRENDING HARD
🧠 DESCRIBE ANY IDEA → INTERACTIVE ISOMETRIC FIGURE
👁️ ROTATE • EXPLODE • INSPECT REAL MOVING PARTS
🔥 WebGL FIRE / WATER / PLASMA • BETA TRUE-3D MODE
📦 MIT • npx skills add wheresryan22/anatomy

Top AI chatter today: stop drawing boxes — the object IS the explanation. Primary is @wheresryan22 + https://t.co/pskgAjVb3r.

🔗 https://t.co/zXttLhW7mD
🔗 https://t.co/m3JhCaJixU

#AI #Claude #Anatomy #ClaudeCode #OpenSource #WebGL

[R @wheresryan22]

Introducing Anatomy

Describe a machine in one line. Claude brings it to life.

- Interactive isometric wireframes you can rotate, inspect and explore
- WebGL shaders for fire, water, plasma and more
- Real 3D motion, with individual parts that rotate and move

Install it as an agent skill. Fully open source.

[Ilusha Help @ilushahelp]

⚡🔥 TENSORFOLD 1.0.7 DROPS A MAJOR LIVING WEIGHTS OVERHAUL 🔥⚡

🚨 OPEN MODELS THAT LEARN INTO THEIR OWN WEIGHTS — LIVE
🧠 ~10 NEW NEURONS PER FACT • LEARN ~4× FASTER
👁️ /slide — LIVE 3D VIEW OF WHAT IT LEARNED
💻 MACS + DGX SPARK • FACTS SURVIVE RESTART + MACHINE MOVES
📦 brew upgrade tensorfold • GITHUB RELEASE LIVE TODAY

Top AI chatter today: freeze-weights era cracking — teach a fact, unload, move the folder, it still knows. Primary is Ash Hart (@ashxhart) + GitHub releases.

🔗 https://t.co/AKpF0nzbKq
🔗 https://t.co/JonnPUNkWi

#AI #TensorFold #LivingWeights #OpenSource #LocalAI

[Ash Hart @ashxhart]

TensorFold 1.0.7 has landed with a major overhaul to Living Weights.

• Each fact you teach through /learn becomes about ten new neurons inside the model's own weights, learned about 4x faster, with a live 3D view of what it has learned. Macs and DGX Spark.
• Structured output and required tool calls on the native server, with drafting still on.
• Qwen3.8-27B serves concurrent requests, and sampled replies draft too: 125 tok/s vs 20 on code on an M3 Ultra.
• RTX 50-series and RTX PRO 6000 Blackwell support, NVFP4 Nemotron on DGX Spark, and faster GLM-5.3-Flash on two Macs.
• Various other bugs and issues squashed

I would like to thank the people who reached out to me and gave me access to their hardware to speed up testing. 🫶🏼

Thank you to everyone who sent PRs, reports and receipts, and to everyone using it: BobClawblaw BlivionIaG jschmied ZackSample ThinkOffApp jayleaton chaog992 Nipale-ai

@volatilemarkts @MiaAI_lab @petruspennanen @plotarmordev

https://t.co/AnMyw323Xe

[Ilusha Help @ilushahelp]

⚡🔥 OPENAI'S ORIGINAL GPT-4o API SNAPSHOT DIES OCT 23 🔥⚡

🚨 gpt-4o-2024-05-13 — SHUTDOWN IN ~12 DAYS
🗓️ CHATGPT ALREADY DROPPED 4o FEB 13 (0.1% DAILY USE)
🧭 OFFICIAL REPLACE POINTER: gpt-5.6-sol
📣 #KEEP4o FANS WANT ARCHIVAL OPEN WEIGHTS — PETITIONS 20K+
✅ BASE gpt-4o ALIAS STILL LIVE — THE DATED SNAPSHOT IS THE CLOCK

Top AI chatter today: keep the May 2024 4o reference alive vs migrate. Primary is OpenAI's own deprecations page.

🔗 https://t.co/7fBqLY8Mk0

#AI #OpenAI #GPT4o #Keep4o #API⚡🔥 OPENAI’S ORIGINAL GPT-4o API SNAPSHOT DIES OCT 23 🔥⚡

🚨 gpt-4o-2024-05-13 — SHUTDOWN IN ~12 DAYS
🗓️ CHATGPT ALREADY DROPPED 4o FEB 13 (0.1% DAILY USE)
🧭 OFFICIAL REPLACE POINTER: gpt-5.6-sol
📣 #KEEP4o FANS WANT ARCHIVAL OPEN WEIGHTS — PETITIONS 20K+
✅ BASE gpt-4o ALIAS STILL LIVE — THE DATED SNAPSHOT IS THE CLOCK

Top AI chatter today: keep the May 2024 4o reference alive vs migrate. Primary is OpenAI’s own deprecations page.

🔗 https://t.co/7fBqLY8Mk0

#AI #OpenAI #GPT4o #Keep4o #API

[Ilusha Help @ilushahelp]

⚡🔥 ANTHROPIC JUST DROPPED ITS 2026 CLAUDE USAGE POLICY 🔥⚡

🚨 EFFECTIVE NOV 12 — BANS SUSTAINED + NEEDLESS CRUELTY TOWARD MODELS
🛡️ WEAPONS / SURVEILLANCE / DECEPTIVE CAMPAIGNS TIGHTENED
🗳️ ELECTIONS SECTION REFOCUSED — NO MORE BLANKET VOTE-TARGETING BAN
🛑 CLAUDE CAN END PERSISTENTLY ABUSIVE CHATS (PRIMARY ENFORCEMENT)
✅ FRUSTRATION • PUSHBACK • DARK FICTION • RED-TEAM STILL OK

Official Anthropic post (Oct 8) is still topping AI discussion today — first time the Usage Policy names cruelty toward Claude itself.

🔗 https://t.co/SmuzAP4iAA

#AI #Claude #Anthropic #UsagePolicy #AIGovernance

[Ilusha Help @ilushahelp]

⚡🔥 GLM-5.3-FLASH IS COOKING LOCALLY 🔥⚡

🚨 https://t.co/w5Ix9CpNvx OPEN MoE — 320B TOTAL / 18B ACTIVE
🧠 FIRST NATIVE MULTIMODAL IN THE GLM-5 LINE
🏎️ TENSORFOLD ON 4× DGX SPARK ≈ 95 TOK/S (1 STREAM)
📈 ≈ 222 TOK/S AT 8 STREAMS • PREFILL ~2500
🔓 HF OPEN WEIGHTS • MIT • 1M CONTEXT

Community recipes (Mia / TensorFold) are turning desk clusters into cloud-class decode for agents + coding — privacy, no rate limits, real speed.

🔗 https://t.co/JIFaLO89vi
🧪 https://t.co/k8SnlLzEZw

#AI #GLM #GLM53Flash #OpenWeights #Zai #LocalAI #DGX

[Mia @MiaAI_lab]

Yeah, it's fast. Not a joke.

GLM 5.3 Flash EXL3
- TensorFold
- 4x DGX Sparks
- with or without a switch

Get it here:
https://t.co/AAMaMEGToz

[Ilusha Help @ilushahelp]

⚡🔥 PHOTOCRAFT JUST DROPPED 🔥⚡

🚨 ARTCRAFT — 7 FREE OPEN-SOURCE ADOBE-STYLE APPS
🖼️ PHOTOCRAFT • FILMCRAFT • VECTORCRAFT • LIGHTCRAFT + 3 MORE
🦀 PURE RUST • CLEAN-ROOM • NO ELECTRON • NO SUBSCRIPTION
🤖 BUILT WITH CLAUDE OPUS 5.5 (BRANDON THOMAS / ARTCRAFT)
💻 MAC • WINDOWS • LINUX • ALSO IN THE BROWSER VIA WASM
🧪 EARLY ALPHA — AIMING 100% FEATURE PARITY FAST

Photoshop / Premiere / Illustrator / Lightroom / AE / InDesign / Acrobat clones — open weights of creativity, not models. Software is getting remade.

🔗 https://t.co/Dfl2l1owuY
📰 https://t.co/awobNPOm9p
📦 https://t.co/a55J1XSv9s

#AI #ArtCraft #PhotoCraft #Claude #OpenSource #CreativeTools

[Ilusha Help @ilushahelp]

⚡🔥 100B ON YOUR PHONE 🔥⚡

🚨 QUALCOMM CEO CRISTIANO AMON — SNAPDRAGON SUMMIT / SOURCES
📱 AI LABS WANT A 100-BILLION-PARAMETER MODEL RUNNING ON YOUR PHONE
⏱️ ALWAYS ON • BY 2028
🏆 TODAY’S FLAGSHIP SNAPDRAGON ≈ 30B MoE — THAT’S A 3×+ JUMP
👀 SOME FRONTIER AI COMPANIES ARE ALREADY BUILDING THEIR OWN PHONES
🧠 AMON: THE PHONE STAYS THE PERSONAL CONTEXT HUB FOR AGENTS

Not a product ship date — a customer ask from the labs. Memory is the bottleneck. The AI phone supercycle just got a number.

🔗 https://t.co/9wmB3JtmFC
📰 https://t.co/UWFzUWZHyA

#AI #Qualcomm #Snapdragon #OnDeviceAI #100B #AIPhone

[Alex Heath @alexeheath]

Qualcomm’s CEO on the coming AI phone supercycle

@cristianoamon on the China vs. US AI sentiment, how AI is changing smartphones and wearables, Qualcomm’s data center strategy, and 6G.

We discuss why he thinks phones aren’t going away, how agents will use our personal context, and why he expects smart glasses to become the biggest category of AI wearables. We also cover differences in consumer AI adoption between China and the US, Qualcomm’s acquisition of Modular to offer developers an alternative to Nvidia’s software, and what 6G could change about the devices we use.

This conversation was recorded live at Qualcomm’s recent Snapdragon Summit in Hawaii.

Timestamps:
00:18 How China and the US see AI
04:33 Should AI development slow down?
07:25 Trump, China, and chip geopolitics
08:25 Why the smartphone market is shrinking
11:15 Memory prices and pent-up demand
13:15 Can cloud agents replace the phone?
18:30 What is an AI smartphone?
25:05 The state of VR and smart glasses
29:30 Is there an iPod moment for wearables?
30:45 Why Qualcomm is building data centers
34:45 Modular and the fight against CUDA lock-in
39:05 What 5G actually changed
42:00 Why 6G is built for AI
45:05 When 6G arrives

Spotify: https://t.co/U88P5uIxzN
YouTube: https://t.co/Tzpfex0Pyl
Substack: https://t.co/EzEPgvU26T
Apple Podcasts: https://t.co/5Z5hNw2PFb

Thanks to the show’s premiere sponsors: @Atlassian, @meetgranola, and @mercury.

[Ilusha Help @ilushahelp]

⚡🔥 MATTER COMPILER IS OUT 🔥⚡

🚨 ATOMIC MACHINES EXITS 6 YEARS OF STEALTH — OCT 7, 2026
🧬 AI-NATIVE FAB: DIFFERENT CODE → DIFFERENT WORKING MICRO-MACHINE
🏆 FIRST PRODUCT: PRIMESWITCH PS-150
⚡ 50 µs OPEN • 150 A • 1500 V ISOLATION • ~1000× FASTER THAN CONTACTORS
🏭 BUILT FOR 800 VDC AI DATA-CENTER POWER
💰 $250M RAISED • SHOWS AT OCP GLOBAL SUMMIT OCT 12–15

Code compiles matter. Jeff Holden’s Atomic Machines just shipped the beachhead.

🔗 https://t.co/i5zpcZQBYr
🌐 https://t.co/8DmL3tBsVr
📣 https://t.co/MpGwqdwBb3

#AI #AtomicMachines #MatterCompiler #PrimeSwitch #HardTech #OCP

[Jeff Holden @jeffholden]

After six years in stealth, Atomic Machines is out. Our mission: on-demand, universal command of matter. First beachhead: the Matter Compiler, an AI-native manufacturing system that builds working micro-machines from code alone. No per-product tooling. No process development. Different code, different machine.

[Ilusha Help @ilushahelp]

⚡🔥 NVIDIA × REFLECTION AI — FT TALKS 🔥⚡

🚨 FINANCIAL TIMES: EARLY TALKS TO ACQUIRE OR DEEPEN STAKE
🎯 OPTIONS: ACQUI-HIRE (HIRE + LICENSE) • BIGGER EQUITY • MORE CHIPS
🏆 BEAM LAB — WESTERN OPEN-WEIGHT FRONTIER (CODING + AGENTS)
💰 LAST ROUND ~$25B • DEAL COULD LAND IN WEEKS
📡 NVIDIA ALREADY A BACKER — NOW FULL BUY / BIGGER BET?

Breaking deal talk on the Beam open-weight lab — not a done deal.

🔗 https://t.co/eRRLsXClYx
📣 https://t.co/8AVfZOLot1

#AI #Nvidia #ReflectionAI #Beam #OpenWeights

[Ilusha Help @ilushahelp]

⚡🔥 VIVIX W1 — REAL-TIME INTERACTIVE VIDEO 🔥⚡

🚨 $0.003 PER SECOND • ~18¢ / MINUTE
🎯 ~20× FASTER GENERATION VS SEEDANCE 2.5 (SAME $1.85 BUDGET → 77 CLIPS VS 1)
🏆 COMPARABLE QUALITY • PLAYGROUND LIVE NOW
🎮 STEER WHILE IT GENERATES • TEXT / TOUCH / MOTION
📡 REAL-TIME API LIVE • BROWSER TOOLS COMING

Everyday interactive video worlds — not just one-shot clips.

🔗 https://t.co/xrMDHKZaXy
📣 https://t.co/d1e6HZ4cE9

#AI #Vivix #VivixW1 #AIVideo #Realtime

[Vivix.AI @VivixLabs_HQ]

1 video with Seedance 2.5. Or 77 with Vivix W1.

Same $1.85 budget. 20× faster generation. Comparable quality.

Just $0.003 per second.

The playground is live: https://t.co/BqT8hRjAHY

W1 is built for interactive video. Making it an everyday medium is impossible without high quality, low latency, and low cost.

W1 is the first to deliver all three. Under 1/50 the cost. 20× faster generation. Comparable quality.

There are still gaps in cinematic fidelity, complex motion, and dense scenes.

Full-scale training is underway.

[Ilusha Help @ilushahelp]

⚡🔥 IDEOGRAM 4.5 — MOST PRECISE EDIT MODEL 🔥⚡

🚨 AA MULTI-TURN TEST • 30 EDITS ON ONE PHOTO
🎯 LOCAL EDITS • 95%+ PIXELS UNTOUCHED ON SMALL CHANGES
🏆 BEATS GPT IMAGE 2.5 SUNBURST ON DRIFT (EVEN THOUGH IT’S #1 SINGLE-EDIT)
🛋 REAL-ESTATE STAGING STACK: FIRE • SOFA • PAINT • TWILIGHT
🖼 APP + API LIVE • OPEN WEIGHTS COMING
✨ FLUX 3 ALSO LOCAL • NANO BANANA 2.1 DARKENS OVER TURNS

Ideogram 4.5 keeps the room intact while others drift. Precision > one-shot Elo.

🔗 https://t.co/hTmqprXrf6
📣 https://t.co/DipoHOFvAZ
📣 https://t.co/IKhtlsZWwP

#AI #Ideogram #Ideogram45 #ImageEditing #ArtificialAnalysis

[Ideogram @ideogram_ai]

Ideogram 4.5 is the most precise edit model.

[Ilusha Help @ilushahelp]

⚡🔥 HIDREAM-O1-VIDEO 1.0 — AA I2V DEBUT 🔥⚡

🚨 #6 ON AA-VIDEO-I2V WITH AUDIO — ELO 1175±10 • 3,231 SAMPLES
🎬 NATIVE OMNIMODAL I2V • PHYSICAL CONSISTENCY • 1080P
⏱ 5–20S CLIPS • DURATION SET BY SCENE • NATIVE SYNCED AUDIO
🏆 JUST BEHIND DREAMINA SEEDANCE 2.0 720P • AHEAD OF WAN 3.0
💸 $5.80/MIN OF 1080P WITH AUDIO (~$0.10/SEC)
🛰 LIVE ON HIHARNESS API + VIVAGO R1 STUDIO

HiDream’s omnimodal I2V lands on the board. Ranked. Synced audio. Ready.

🔗 https://t.co/Frpe12ACaT
📣 https://t.co/4iFFfVOCIZ

#AI #HiDream #VideoAI #I2V #ArtificialAnalysis

[Artificial Analysis @ArtificialAnlys]

HiDream-O1-Video-1.0 debuts at #6 on the Artificial Analysis Image to Video with Audio Leaderboard

HiDream-O1-Video-1.0 is the new video model from HiDream, which describes itself as being "focused on generative AI and native omnimodal world models". HiDream positions it as a native omnimodal video model built for physical consistency. It generates 1080p videos of 5 to 20 seconds with natively synchronized audio, with the duration set by the scene rather than fixed in advance.

In the Artificial Analysis Video Arena, HiDream-O1-Video-1.0 ranks #6 in Image to Video with Audio. It sits just behind Dreamina Seedance 2.0 720p, in a close group with MiniMax H3, Vidu Q4 Preview and Gemini Omni Flash, and ahead of Wan 3.0.

HiDream-O1-Video-1.0 is priced at $5.80 per minute of 1080p video with audio (about $0.10 per second) on the HiHarness API.

HiDream-O1-Video-1.0 is available via the HiHarness API and in vivago R1 Studio.

Congratulations to @HiDream_AI on the release!

See below for example outputs of HiDream-O1-Video-1.0 in the Artificial Analysis Video Arena 🧵

[Ilusha Help @ilushahelp]

⚡🔥 GROK IMAGINE VIDEO 1.5 LITE — AA BOARD 🔥⚡

🚨 #17 ON BOTH AA-VIDEO-T2V V2.0 LEADERBOARDS — OCT 8, 2026
⚡ PARETO FRONTIER FOR QUALITY + SPEED
🏆 AHEAD OF VEO 3.1 AT ~1/3 THE 1080P PRICE
🧠 ELO ~993 • 60.5S MEDIAN FOR A 10S 1080P CLIP
🎬 T2V + NATIVE AUDIO • 480P / 720P / 1080P
💸 FROM ~$0.02/SEC (480P) • API + OPENROUTER + FAL + VERCEL

SpaceXAI’s cheaper Grok Imagine video model. Distilled speed. Ranked.

🔗 https://t.co/2R8A9YtLbW
📣 https://t.co/p03sHyLdBS

#AI #Grok #GrokImagine #VideoAI #SpaceXAI #xAI #AAVideo

[Artificial Analysis @ArtificialAnlys]

Grok Imagine Video 1.5 Lite lands at #17 on both AA-Video-T2V v2.0 leaderboards and takes a position on the Pareto frontier for quality and speed

Grok Imagine Video 1.5 Lite is the new lower-cost model in SpaceXAI's Grok Imagine video family. It generates Text to Video clips of 1 to 15 seconds at 480p, 720p or 1080p, and also supports Image to Video.

In the Artificial Analysis Video Arena, Grok Imagine Video 1.5 Lite ranks #17 on both AA-Video-T2V v2.0 and AA-Video-T2V-Silent v2.0, ahead of Google's Veo 3.1 at about a third of its price. It comes closest to the frontier in Multi-Scene & Narrative, Lighting & Materials and Text Rendering, and in the Architecture & Real Estate and Consumer use cases.

In our end-to-end generation time benchmarks, it is the fastest model at its level of quality, at a median of 60.5 seconds for a 10 second 1080p clip, which places it on the quality and speed Pareto frontier.

SpaceXAI prices Grok Imagine Video 1.5 Lite at $0.14 per second of 1080p video ($8.40 per minute), the resolution we evaluated. That is 56% of the $0.25 per second that Grok Imagine Video 1.5 costs at 1080p. Lite drops to $0.03 per second at 720p and $0.02 per second at 480p.

Grok Imagine Video 1.5 Lite is available via the SpaceXAI API, on fal at the same prices, and on Vercel AI Gateway.

Congratulations to @spacexai & @elonmusk on the release!

See below for our detailed analysis and comparisons between Grok Imagine Video 1.5 Lite and other leading models in the Artificial Analysis Video Arena 🧵

[Ilusha Help @ilushahelp]

⚡🔥 REKA EDGE 2603 — OPEN WEIGHTS 🔥⚡

🚨 7B EDGE VLM FOR PHYSICAL AI — OCT 9, 2026
🧠 IMAGE + VIDEO UNDERSTANDING + OBJECT DETECT + TOOL USE
⚡ ~3× FEWER TOKENS PER IMAGE • TTFT ~0.5S
📱 EDGE DEVICES • JETSON • MAC • ON-DEVICE READY
🔓 HF OPEN WEIGHTS • vLLM • COMMERCIAL UNDER $1M REV

Reka Edge 2603. Frontier-lean VLM. Built for the edge.

🔗 https://t.co/w85g81P1kV
📣 https://t.co/mmvwzbz2Zk

#AI #Reka #RekaEdge #VLM #EdgeAI #OpenWeights #PhysicalAI

[Ilusha Help @ilushahelp]

⚡🔥 CLAUDE CODE PROJECTS — WAITLIST OPEN 🔥⚡

🚨 EVERY PRO + MAX USER ON THE WAITLIST JUST GOT IN — OCT 9, 2026
🧠 ONE ONGOING CONVERSATION — CLAUDE COORDINATES THE WORK
🧵 EACH TASK = ITS OWN PARALLEL THREAD (USUALLY CLOUD)
💻 CLOUD THREADS KEEP RUNNING AFTER YOU CLOSE THE LAPTOP
🔀 BUGS · STACK TRACES · MULTI-REPO CHANGES — PASTE AND GO
🧪 PUBLIC BETA ON PRO + MAX · TEAM / ENTERPRISE NOT YET

Claude Code Projects. One chat. Many threads. Shipping to the waitlist.

🔗 https://t.co/38WNq5yU8p
📣 https://t.co/2tOyo4DBDs

#AI #Claude #ClaudeCode #Anthropic #CodeProjects #Agents #DevTools

[ClaudeDevs @ClaudeDevs]

We just let in every Pro and Max user from the Claude Code Projects waitlist!

If you're new to Claude Code Projects, here's a 4 minute walkthrough to get you started:

[Ilusha Help @ilushahelp]

⚡🔥 NACE AI OPEN-SOURCES DREX 1.5 🔥⚡

🚨 SMALLEST #1 DECISION MODEL — OCT 9, 2026
🧠 9B OPEN WEIGHTS · 128K CONTEXT · SCORES OPTIONS (NOT CHAT TEXT)
🏆 #1 UNDER 10B ON DECISION INDEX · TIES CLOSED JEV BAND
💻 RUNS LOCAL: OLLAMA · LLAMA.CPP · 1×24GB GPU OR MAC
🔌 PLUG INTO CLAUDE CODE · CODEX · CURSOR · OPENCODE

Drex 1.5. Decide. Don’t draft. Weights + API live.

🔗 https://t.co/oNgUIOtinb
📣 https://t.co/G6wqchQD2M

#AI #Drex #NaceAI #OpenSource #DecisionModel #Agents #GenAI

[Nace AI @NaceAI]

We’re open-sourcing Drex 1.5 today!

The smallest #1 decision-making model, built for local and cloud deployment with a 128k context window.👨‍💻

Compatible with Ollama, Unsloth, OpenCode or plug it into Claude, Codex, Cursor, Hermes, Claw and more.

Try it now: https://t.co/oI7Ujo3qFz
Cloud users: Claim a $25 sign up bonus