Apolyst
108 posts
Apolyst 리트윗함
Yeahh Meta back at it! 🔥
Alexandr Wang@alexandr_wang
new research from Meta FAIR: Code World Model (CWM), a 32B research model we encourage the research community to research this open-weight model! pass@1 evals, for the curious: 65.8 % on SWE-bench Verified 68.6 % on LiveCodeBench 96.6 % on Math-500 76.0 % on AIME 2024 🧵
English
Apolyst 리트윗함
Apolyst 리트윗함
There's another new stealth model in Cursor from one of our partners.
"code-supernova" is free to try, curious how you find it!
English
Apolyst 리트윗함
🎁 GPT-5 is 50% off on OpenRouter - for a week
- No opt-in needed 🚀
- Discount limits to 20 RPM, standard price after that
- From Sept 17th, 10 AM PST to September 24th, 10 AM PST
Give @OpenAI GPT-5 a try openrouter.ai/openai/gpt-5
English
Apolyst 리트윗함
We trained gpt-5-codex to be great at both responsive and mobile front-ends. Here's a thread of some examples:
"Make a pixel art game where I can walk around and talk to other villagers, and catch wild bugs."
English
I thought we'd overcome these hallucinations. Double bug in the Gemini app using Nano Banana. It went bananas indeed. #NanoBanana #Gemini #bug
English
Apolyst 리트윗함
Apolyst 리트윗함
Introducing VEED Fabric 1.0
The world’s first-ever AI talking video model.
– 1-minute-long videos
– 60x cheaper
– 7x faster
🚨 Giving away unlimited FREE credits for the next 24 hours.
(until they run out).
Want access?
• Retweet
• Leave a comment
English
Apolyst 리트윗함
🚀 Introducing Qwen3-Next-80B-A3B — the FUTURE of efficient LLMs is here!
🔹 80B params, but only 3B activated per token → 10x cheaper training, 10x faster inference than Qwen3-32B.(esp. @ 32K+ context!)
🔹Hybrid Architecture: Gated DeltaNet + Gated Attention → best of speed & recall
🔹 Ultra-sparse MoE: 512 experts, 10 routed + 1 shared
🔹 Multi-Token Prediction → turbo-charged speculative decoding
🔹 Beats Qwen3-32B in perf, rivals Qwen3-235B in reasoning & long-context
🧠 Qwen3-Next-80B-A3B-Instruct approaches our 235B flagship.
🧠 Qwen3-Next-80B-A3B-Thinking outperforms Gemini-2.5-Flash-Thinking.
Try it now: chat.qwen.ai
Blog: qwen.ai/blog?id=4074cc…
Huggingface: huggingface.co/collections/Qw…
ModelScope: modelscope.cn/collections/Qw…
Kaggle: kaggle.com/models/qwen-lm…
Alibaba Cloud API: #c5414da58bjgj" target="_blank" rel="nofollow noopener">alibabacloud.com/help/en/model-…
English
For simple tasks balancing thinking and code: claude-4-sonnet
----
This works for me. If anyone has better options, please share! #VibeCoding
English
Current vibe coding workflow:
High-level ideas w&w/o code execution: gpt-5-high
Start & major revamps with high-level code: gpt-5-high
Code execution with clear instructions: grok-code-fast-1
Complex tasks needing both thinking and code execution, if credits: claude-4.1-opus
>>
English
Apolyst 리트윗함
We’ve (finally) added full support for MCP tools in ChatGPT.
In developer mode, developers can create connectors and use them in chat for write actions (not just search/fetch). Update Jira tickets, trigger Zapier workflows, or combine connectors for complex automations.
English
English
