HNHacker News
TopNewBestAskShowJobs

vinhnx

9,121 karma · joined May 16, 2012

[ my public key: https://keybase.io/vinhnx; my proof: https://keybase.io/vinhnx/sigs/BL88w_oUlDAOOxOvQLzYVvQBWsBp-5bouAYj3xCBC4I ]

Hi, I'm Vinh (@vinhnx on internet). iOS & Applied AI Engineer. I love to learn, build, and share.

Currently I'm building VT Code. VT Code is an open-source Rust terminal coding agent. https://github.com/vinhnx/VTCode

If you like my sharing. You can buymeacoffee via https://buymeacoffee.com/vinhnx

Thank you for your support!

Twitter: https://x.com/vinhnx GitHub: https://github.com/vinhnx HuggingFace: https://huggingface.co/vinhnx90 Personal site: https://vinhnx.github.io CV: https://vinhnx.github.io/cv YouTube: https://www.youtube.com/vinhnx90 Buy me a coffee: https://buymeacoffee.com/vinhnx

Learn by doing.

submissionscomments
vinhnx··on DeepSeek v4
The king is back! I remember vividly being very amazed and having a deep appreciation reading DeepSeek's reasoning on Chat.DeepSeek.com, even before the DeepSeek moment in January later that year. I can't quite remember the date, but it's the most profound moment I have ever had. After OpenAI O1, no other model has “reasoning” capability yet. And DeepSeek opens the full trace for us. Seeing DeepSeek's “wait, aha…” moments is something hard to describe. I learned strategy and reasoning skills for myself also. I am always rooting for them.
vinhnx··on Claude Code to be removed from Anthropic's Pro plan?
Anthropic reverts the “prosumer” AB testing, but damage is done.
vinhnx··on A Brief History of Fish Sauce
Oh absolutely and you're welcome! Btw, fish sauce in scrambled eggs over rice is one of the simplest, most satisfying meals you'll find across Southeast Asia, in my country Vietnam especially. It's my favorite meal also.
vinhnx··on Codex for almost everything
A simple mental model for Claude's new adaptive thinking is that it is the recommended way to use extended thinking. Adaptive Thinking (wraps Extended Thinking). It applies to Opus 4.7, 4.6, and Sonnet 4.6 and is the default mode on Claude Mythos Preview.
vinhnx··on Claude Opus 4.7
They do now. /effort command is on the latest Claude Code version; run `claude update` and `claude /effort`.
vinhnx··on The Image Boards of Hayao Miyazaki
You're welcome! I shared from AnimationObsessive's blog, and they did a great deal of research and showed the drawing image boards on paper from Mr. Hayao Miyazaki himself and the studio. This article I find unique, and some images have not been seen or public before. I recommend subscribing to their blog for more animation articles.
vinhnx··on Mathematics Is All You Need
I'm sorry.
vinhnx··on What about Juniors?
Thank you!
vinhnx··on Mathematics Is All You Need
Sorry for the confusion, even though the author's names may not have an active record on Scholar. But I would like to share it here because I read the paper, and I find it interesting.
vinhnx··on Before You Adapt to AI… Watch This
Sorry, I saw the post-credit title and I thought this was the Author's intended title. I changed it back now.
vinhnx··on LLM Architecture Gallery
You're welcome!
vinhnx··on OpenCode – Open source AI coding agent
I've been building VT Code (https://github.com/vinhnx/vtcode), a Rust-based semantic coding agent. Just landed Codex OAuth with PKCE exchange, credentials go into the system keyring.

I build VT Code with Tree-sitter for semantic understanding and OS-native sandboxing. It's still early but I confident it usable. I hope you'll give it a try.

vinhnx··on Astral to Join OpenAI
What excites me about the OpenAI + Astral acquisition: Codex CLI, uv, and ruff are all written in Rust. Fast by design, and fully open source.
vinhnx··on JPEG Compression
I think my submission about this post was selected to "second-chance" pool by HN Moderators. Hence it's being shown again. Thanks for the heads up!
vinhnx··on LLM Architecture Gallery
I think Cognition DeepWiki's or Google CodeWiki's code map does generated a architecture map (Mermaid style). Eg: https://deepwiki.com/openai/codex#project-purpose-and-archit...
vinhnx··on Ask HN: What Are You Working On? (March 2026)
This month, I'm working on VT Code, a terminal-native coding agent I've been building in Rust (https://github.com/vinhnx/vtcode).

This month I'm focusing on long-pending TODO items: self-benchmarking with Terminal bench (https://www.tbench.ai/), fuzzing the security parsers (it executes shell commands, so the threat model is real), normalizing extended thinking traces across providers, and improving the agent UI/UX and TUI components and harness.

vinhnx··on What podcasts are you listening to?
I listen to a lots of podcast.

Currently here is my pinned favorites ones:

Practical AI

Grit

Wenbin Fang's Podcast Playlist (Founder of https://www.listennotes.com/)

The Gradient: Perspectives on AI

Latent Space: The AI Engineer Podcast

The Empty Bow

Hacker News Recap

Dwarkesh Podcast

Machine Learning Street Talk (MLST)

Interconnects

Talk Python To Me

Training Data

---

My podcast collection (opml file format). Exported from Overcast.

Feel free to import to your podcast app of choice. https://github.com/vinhnx/podcasts

vinhnx··on Mercury 2: Fast reasoning LLM powered by diffusion
This research paper "Mercury: Ultra-Fast Language Models Based on Diffusion" from last year (2025)

https://arxiv.org/pdf/2506.17298

vinhnx··on Fix your tools
You're welcome! Love the article, I hope you write more.
vinhnx··on Gemini 3.1 Pro
Model card https://deepmind.google/models/model-cards/gemini-3-1-pro/
vinhnx··on Ask HN: What are you working on? (February 2026)
This month I'm currently continuing to work on my opensource coding agent VT Code (https://github.com/vinhnx/VTCode). For the last few months I've been improving the harness and UI/UX. Focusing developer experience and TUI performance.
vinhnx··on Claude Opus 4.6
VS Code confirms that they are experimenting with the new adaptive thinking and high reasoning effort params. https://x.com/pierceboggan/status/2019645801769689486
vinhnx··on Claude Opus 4.6
Confirm by PM lead at VS Code team

> "We have high thinking as default + adaptive thinking, first time we’ve run with these settings..."

> https://x.com/pierceboggan/status/2019645801769689486

vinhnx··on Claude Opus 4.6
Just used Opus 4.6 via GitHub Copilot. It feels very different. Inference seems slow for now. I guess Opus 4.6 has adaptive thinking activated by default.
vinhnx··on The Codex App
Best part of the Codex app launch is that, OpenAI has opened the whole Codex ecosystem (CLI, Web, IDE Extensions) for free ChatGPT users. And x2 usage for Plus, Pro. This is I think to gain developers' attraction from Claude Code.
vinhnx··on A few random notes from Claude coding quite a bit last few weeks
Boris Cherny (Claude Code creator) replies to Andrej Karpathy

https://xcancel.com/bcherny/status/2015979257038831967

vinhnx··on Kimi Released Kimi K2.5, Open-Source Visual SOTA-Agentic Model
It is, Kimi Code CLI supports Zed' Agent Client Protocol (http://agentclientprotocol.com/), so it can acts as an external agent that could run in any ACP-compatible client, eg: Zed, Jetbrain, Toad CLI, Minano Notebook. Also, it supports Agent Skills. Moonshot AI developers are actively update the agent and every active. I really like their CLI.
vinhnx··on Kimi Released Kimi K2.5, Open-Source Visual SOTA-Agentic Model
One thing caught my eyes is that besides K2.5 model, Moonshot AI also launched Kimi Code (https://www.kimi.com/code), evolved from Kimi CLI. It is a terminal coding agent, I've been used it last month with Kimi subscription, it is capable agent with stable harness.

GitHub: https://github.com/MoonshotAI/kimi-cli

vinhnx··on Unrolling the Codex agent loop
This came as a big surprise to me last year. I remember they announced that Codex CLI is opensource, and the codex-rs [0] from TypeScript to Rust, with the entire CLI now open source. This is a big deal and very useful for anyone wanting to learn how coding agents work, especially coming from a major lab like OpenAI. I've also contributed some improvements to their CLI a while ago and have been following their releases and PRs to broaden my knowledge.

[0] https://github.com/openai/codex/tree/main/codex-rs

vinhnx··on Ask HN: What are you working on? (January 2026)
This January, I'm still working on my coding agent VT Code. Hardening execution harness, improving tool discovery mechanisms, refining context engineering and performance optimization.

https://github.com/vinhnx/vtcode

← PreviousPage 3 of 8Next →