Register and share your invite link to earn from video plays and referrals.

NZ
@CodeByNZ
AI Insider / Tracking the race to AGI Startups • Agents • Open Source Sharing what matters before it goes mainstream
752 Following    18.9K Followers
Qwen4-Max just dropped, so I had it make me a motion reel It was benchmaxxed, but the design taste is great Pure code, zero plugins, every frame written in code Prompt: make a dynamic 15-second motion graphics video that shows what an incredible motion designer you are, like it's your showreel for a résumé. go all out
Show more
Rumors say Anthropic's mysterious Model 4 is actually Fable 5.5. And Model 5 is supposedly already in training for harder math, biology and science, and could end up behind both Fable 6 and Opus 6. Nothing is confirmed yet. Still, OpenAI should be VERY worried right now.
Show more
$26.54 per task for fable 5.2. $0.29 for gemini 4 hydrogen. that's about 91x more on artificial analysis' numbers. opus 5.5 is around $10, gpt 6.1 sol is $1.27, and hydrogen still undercuts all of them. fable is still the one for hard stuff, but if you're running a lot of tasks, hydrogen is the cheap pick right now.
Show more
Gemini 4 Oxygen, reportedly the next Flash model, and the design taste is really good. train test (The Copper Kestrel), 1 shot in around 5 minutes: - assemble and explode views with a spread slider - wireframe, x-ray, drive gear, inspect and labels - 7 camera angles: hero 3/4, elevation, front, footplate, chase, aerial loop and auto camera
Show more
OpenAI might have its strongest models ever sitting in the lab, and none of them can ship. The word going around is that 6.1, 6.2 and 6.5 all keep failing safety and alignment checks, and the delays are already weeks long. Meanwhile Anthropic shipped Opus 5.5 and Fable 5.2 with no drama. If 6.2 drops soon, expect a lot of guardrails on day one.
Show more
👀 A leaked Claude Code project settings screen, codenamed Conway, shows separate model pickers for the Coordinator and the Threads. Fable 5.2 or Opus 5.5 can do the planning while a cheaper, faster model handles the worker threads. That alone could save a lot of usage. There's also a toggle to continue threads automatically when your usage limit resets. No more coming back to a paused run. 🙏 Nothing is announced yet, so treat it as a leak for now.
Show more
👀 A leaked Claude Code project settings screen, codenamed Conway, shows separate model pickers for the Coordinator and the Threads. Fable 5.2 or Opus 5.5 can do the planning while a cheaper, faster model handles the worker threads. That alone could save a lot of usage. There's also a toggle to continue threads automatically when your usage limit resets. No more coming back to a paused run. 🙏 Nothing is announced yet, so treat it as a leak for now.
Show more
Qwen4-Max by @Alibaba_Qwen has landed as #6# in the Code Arena: WebDev! With 1750 pts (preliminary), it's +74 pts over its predecessor, Qwen3.8-Max (0902) at 1676. It's also the top Chinese model in: - #6# HTML: 1748 pts - #8# Brand & Marketing: 1705 pts - #8# Reference-Based Design: 1770 pts Congrats to the @Alibaba_Qwen team on this release!
Show more
qwen4-max is out and cost per task went from $4.43 to $1.67 lol up to 50% fewer tokens, 1M context, native vision. now give us the 27b
📢 Meet Qwen4-Max, your new AI teammate for work, creativity, and discovery! Qwen4-Max is built on a new MoE architecture with native vision and a 1M-token context window. Highlights: - Outperforms Qwen3.8-Max with up to ~50% fewer tokens, while average cost per task drops over 60% in our internal evaluation. - Advances real-world productivity through greater proficiency with professional tools, stronger long-horizon execution, and higher-quality deliverables. - Handles real tasks, including 3D modeling, game development, manufacturing, finance, law, healthcare, and frontier scientific discovery. - $1.5/M tokens for input, $4.5/M tokens for output, and $0.075/M tokens for implicit caching — all lower than Qwen3.8-Max, with implicit caching down by 70%. The API is live on QwenCloud. You can also try Qwen4-Max on Qwen Studio, QwenWork and Qoder. Start building today! 🚀 - Blog: - API: - Token Plan: More to come! The open weights of smaller Qwen4 models will be released. Stay tuned. 👀
Show more
gpt-6.2 astra might already be getting sold through api resellers lol lots of people say they have access and are running it on the official codex harness. that azure sighting a few days ago could be how it got out, and openai still hasn't said a word
Show more
Krypton, GPT-6.2 Sol, and Opus 5.6 might all drop this week. Google, OpenAI, and Anthropic in the same week.
OpenAI had a security scare a while back, and since then they’ve been so focused on safety that they keep holding models back. Reports say the next big model is already trained, but it’s still not out. At this point they're just giving Anthropic free time to pull ahead.
Show more
ngl all this OpenAI safety stuff coming out today is just annoying me at this point They said frontier RL training resumed about 6 weeks ago, and we still have no sign of GPT-6.2 Astra or Sol If safety and alignment are still the problem now they are cooked
Show more
OpenAI might have its strongest models ever sitting in the lab, and none of them can ship. The word going around is that 6.1, 6.2 and 6.5 all keep failing safety and alignment checks, and the delays are already weeks long. Meanwhile Anthropic shipped Opus 5.5 and Fable 5.2 with no drama. If 6.2 drops soon, expect a lot of guardrails on day one.
Show more
Gemini 4 Oxygen, reportedly the next Flash model, and the design taste is really good. train test (The Copper Kestrel), 1 shot in around 5 minutes: - assemble and explode views with a spread slider - wireframe, x-ray, drive gear, inspect and labels - 7 camera angles: hero 3/4, elevation, front, footplate, chase, aerial loop and auto camera
Show more
PLEASE SWITCH TO DEEPSEEK FLASH OR TERRA. someone ran luna, deepseek flash and terra through 50 puzzles all three cost about the same, but luna got fewer right and took about 3x more steps to finish
Show more
$26.54 per task for fable 5.2. $0.29 for gemini 4 hydrogen. that's about 91x more on artificial analysis' numbers. opus 5.5 is around $10, gpt 6.1 sol is $1.27, and hydrogen still undercuts all of them. fable is still the one for hard stuff, but if you're running a lot of tasks, hydrogen is the cheap pick right now.
Show more
100 billion tokens and you’re still not #1#
NEW: Sam Altman reveals OpenAI’s top “token leader” uses 100,000,000,000 tokens per month — and still isn’t the highest user in the world.
✅ Claude Code ✅ Claude Cowork ✅ Claude Design 🔲 Claude Marketing 🔲 Claude Sales 🔲 Claude Finance 🔲 Claude HR 🔲 Claude Operations 🔲 Claude Customer …
0
413
14.1K
1.4K
Forward to community
pov : When you give your child a laptop, and he opens the terminal before Chrome
0
24
6.9K
332
Forward to community