
⠋ building cool stuff ⠙ 23 ┃product engineer @HeyGen based in Toronto ┃ Built @editfa_st ┃Building @HyperFrames_ by HeyGen
🇪🇸/🇨🇦/🇺🇸
Joined August 2023
- Tweets13.6K
- Following1.5K
- Followers3.8K
- Likes76.7K
Miguel Ángel retweeted
SWE-bench tells you which model writes working code
Code2Video tells you which one makes a video people want to watch
it's now a standalone leaderboard on @ArtificialAnlys
27 models, judged head to head on engagement, prompt intent, composition, temporal consistency and craft
Claude Opus 5.5 still leads
Check out their article here ↓
Introducing Code2Video-AA, our independent implementation of HeyGen’s Code2Video Bench, which evaluates how well models create product and brand marketing videos in code
Code2Video-AA tests models across 114 design briefs, tasking them with turning each product or brand marketing brief into a @HyperFrames_ HTML composition, which we render to video. Model outputs are evaluated through pairwise comparisons with other model outputs across five axes: Engagement, Prompt Intent, Composition, Temporal Consistency, and Craft. We use the outcomes of these comparisons to calculate an Elo rating for each axis, which are then averaged to produce the headline Code2Video-AA Elo. Code2Video-AA was built in partnership with @HeyGen.
Claude Opus 5.5 (max with fallback, 1536) leads Code2Video-AA by 93 points, but costs $3.08 per task, ~10x more than GPT-6.1 Sol (max, 1369), which trails by 167 points.
Key findings:
➤ Claude Opus 5.5 leads on all five judge axes: Models are judged across 5 axes: Engagement, Prompt Intent, Composition, Temporal Consistency, and Craft. Claude Sonnet 5.5 (max with fallback, 1443) and GPT-6 Astra (max, 1427) are statistically tied for second in overall score
➤ Cost per task varies more than 300x, from $0.01 to $3.89: GPT-6.1 Sol (max), fourth, has a Cost per Task of $0.31, about 90% lower than Claude Opus 5.5 (max with fallback) at $3.08
➤ Open weights models trail by at least 483 points: DeepSeek V4.1 Flash (max, 1053), MiMo-V2.6-Flash (1052) and MiMo-V2.6-Pro (1052) are the top open weights models, at $0.01 to $0.08 per task
Code2Video-AA details:
➤ Models are tested on a single response without tools. We render it with HyperFrames, with no network access
➤ A judge (Gemini 4 Argon, high) compares pair of videos across five axes: engagement, prompt intent, composition, temporal consistency and craft, with the final score being calculated from a combination of each individual axis Elo
➤ Over 35,000 verdicts rank 27 models by Elo, with 95% confidence intervals
More results, example videos and how Code2Video-AA works follow in the thread below
Miguel Ángel retweeted
Improved this video with HyperFrames studio and one prompt
give your features the launch they deserve with @HyperFrames_
Miguel Ángel retweeted
Kurzgesagt has 25M subs. Millions of views on EVERY single video 🤯
nobody copies them for a reason: one video takes 1,200 hours and a studio of 50+ animators.
we rebuilt their format with Opus 5.5 + @HyperFrames_ Studio. 100% code.
a full 11-minute explainer. made in ~2 hours. under $50.
want the template? we’re open-sourcing it for free 👇
Miguel Ángel retweeted
opus 5.5 era changed how code-to-video works
so we deleted ~98k lines of skills from @HyperFrames_ 🤯
smart models don’t need a workflow. they need the best tool to make video
21 skills → 7
470 KB of instructions → 39 KB
and it got way better:
→ agent uses the tools right: 13% → 79%
→ fully correct videos: 11% → 71%
→ 68% cheaper, 76% faster per run
we had to remove 100k LOC of skills in the @HyperFrames_ repo thanks to Opus 5.5
now we have only 7 skills instead of +20 with better quality and speed (from init to finished video) thanks to the herculean efforts from @wenbozhu0328 and @WaterrrMiao!
Miguel Ángel retweeted
Here’s how to use Claude to copy any video shorts style in a few minutes.
From @jake11moran:
“We added this as a native capability in @HyperFrames_ studio. You just upload or pick a reference clip and then describe in the prompt what you want to change.
You can see where we’re drawing inspiration from and apply a different font or different colors. It’s still structurally the same idea.”
📌 Watch our full AI video editing tutorial here: youtu.be/MC1g2fHYC88
Anyone can now make amazing videos with Claude Opus if you know how to prompt it.
Here's my new episode with @jake11moran where he showed me exactly how to use Claude and @HyperFrames_ to:
→ Make a launch video from just a few images
→ Duplicate your favorite creator's video style
→ Make precise edits with the right keywords
Every minute of our 40-min episode is packed with alpha and live demos.
📌 Watch now: youtu.be/MC1g2fHYC88
Thanks to our sponsor:
@WisprFlow: Get 1 month free of the best voice dictation tool wisprflow.ai/?promo_code=PET…
day 2/100 of turning @HyperFrames_ into a real business:
i studied the best oss business models: vercel, databricks, confluent, etc.
sharing my notes on @vercel - an incredible product and a great business built on top of its @nextjs 's massive popularity
the story is fascinating because the business actually came first.
10 years ago TODAY, a small startup called ZEIT open-sourced Next.js. but it was AFTER ZEIT launched Now, a product that made deploying websites as easy as running one command.
Later that year, they released Next.js under the MIT license, completely free for anyone to use. (whoever made this call, is a true genius)
Instead of monetizing the framework, they focused on making it ridiculously easy to take something built with Next.js and deploy it to production.
And it worked.
Next.js became one of the most popular web frameworks. Developers learned it, contributed to it, and built an entire ecosystem around it.
Meanwhile, Vercel (renamed from ZEIT) became the natural next step. Push code, get a live URL, share a preview, collaborate with your team, and scale to production.
Open source drove adoption. The commercial platform captured the value of production.
Here's the crazy part: Vercel grew 50% yoy in 2026 because coding agents defaults to build modern web apps using Next.js
In 2026, Vercel surpassed $500M in annualized revenue run rate, after reaching a $9.3B valuation in 2025.
Little do people know, Vercel has reported that roughly 70% of Next.js applications run outside its platform. Yet, they didn't need to own EVERY Next.js deployment to build a massive business.
To me, the biggest lesson isn't that open source is a great distribution strategy.
It's that the best open-source businesses don't have to restrict what developers can do for free. They create something so valuable around the open-source ecosystem that people naturally want to pay for it.
Free to build. Worth paying to ship.
Incredible execution of technology, product and business by @rauchg and the @vercel team. mad respect.
And happy 10th birthday, @nextjs 🎂🎉
Miguel Ángel retweeted
ICYMI - news from yesterday 🥇 Made with @HyperFrames_ "Reference Any Video" one shot.
HeyGen Video 1 went up against today's leading video models in a blind test on @ArtificialAnlys.
Our first general-purpose video model ranked #1 in text to video.
Try it: platform.heygen.com/models/h…
Miguel Ángel retweeted
Make your videos even stronger with @HyperFrames_ motion graphics.
Same tailored cover letter for every job, now with your stats, charts, and name animated in.
Your cover letter is sitting in a pile with hundreds of PDFs. Send something they can actually watch.
ChatGPT pulls the @indeed listing, and HeyGen MCP turns it into a video cover letter for that exact role, in your face and voice.
Miguel Ángel retweeted
So basically, today PHP got two critical bugs that let anyone run code on your server without even logging in, which is bad, and then the details leaked before the patch was even out, which is worse, and it works on WordPress and MediaWiki and Magento and Joomla and Laravel, and Censys says like 1.5 million servers are running a version with no fix, so the NSA was like "take your system offline," and the FBI was like "take your system offline," and CISA and the UK and CERT-EU were all like "take your system offline," and then the Prime Minister of Japan, the actual Prime Minister, posted at 5:55 in the morning telling 48.8 million people to take their system offline, so everyone took their system offline, like the Dutch cybersecurity agency took its own website offline, and the Dutch parliament went offline, and the Port of Rotterdam went offline, and Finland and France and Denmark started pulling government sites, and Wikipedia went down, and Archive.org went down, and Asana and GOG and Something Awful went down, and the University of Vienna's webmail went down for some reason, and that's just PHP. Because GitLab also dropped a CVSS 10.0, which is the highest score they give out, for a bug that's been there since 2014, so that's every self-hosted GitLab for twelve years, and then LiteLLM 1.107.0 got hijacked and shipped a credential stealer and a remote access trojan in every single wheel and the Docker images, including latest, and you didn't even have to import it, just installing it was enough, and it's the second time this year, and then Railway paused every deployment on the entire platform because someone got into the environment variables, so now everyone has to rotate every secret they've ever stored there, and TiDB Cloud said someone was walking around inside with stolen credentials for basically a month and an extortion group says they took 10.6 terabytes, and Zeabur got breached for the second time in two months and is just shutting down, like the whole company, and Node.js patched remote code execution and a TLS bypass, and CISA says Plex is being actively exploited, and someone took over 300 Tor exit relays to steal SSH passwords, and SafePay ransomware has hit over 600 critical infrastructure orgs, and India's biggest bank filed a data loss incident, and Japan has been in an official state of emergency in cyberspace since October 9, which honestly explains the 5:55 AM post. Oh, and eMudhra, a certificate authority that's issued over 60 million certificates, whose entire job is proving websites are who they say they are, was supposed to be using private keys that are random numbers over a hundred digits long, and one of them was 4. Just 4. And the others were like 66 and 237. Anyway, it's Wednesday.
An urgent notice from the Railway team
blog.railway.com/p/security-…
Miguel Ángel retweeted
Replying to @HyperFrames_
@HyperFrames_ Studio 做出了「一口气了解」的第二集!!!
上一集讲国家,这一集讲有色金属:「一口气了解铜」,8 分 43 秒,又一集她没讲过的题目,从 0 开始。
主持人换成了 @HeyGen 的 Avatar 模型。先别看字,看完视频告诉我:这个数字人,你还分得出来吗?😉
上一条说过,code to video 有三步:复刻一个片段 → 完整复刻 10 分钟 → 创作新的 10 分钟。
泰国那集证明了第三步能做。这一集要回答的是:换一个领域还成不成立。题目、数据、素材、画面全换了,用的还是同一套模板,出来还是那个味道。
复刻有原片可以对照。从 0 开始,每一步都得 agent 自己判断:
- 素材去哪找、找什么
- 每句话配什么画面,怎么卡着文案进出
- 文案怎么写才是小Lin味,而不是流水线式的讲解
- 音效、背景音乐、字幕、数字人怎么和口播对齐
- 声音怎么撑满整集:语序、语调、停顿都要对
@HyperFrames_ Studio 把这些原生解决了:
- 把她拆成一套模板:怎么写、怎么剪、怎么动
- 动效逐帧对齐,1:1 重建
- 音效、音乐、字幕、数字人跟着口播卡点
- HeyGen 的 Avatar 模型 + 专业级声音克隆
国家能做,金属也能做。「一口气了解 XXX」可以一集接一集地往下做。
hyperframes.dev/studio
(视频中的主持人形象和声音均为 AI 生成,非小Lin本人,仅作技术演示)
复刻小Lin说 1 分钟,现在谁都能做。
做满 10 分钟、还是那个味道,是另一回事🤥
Opus 5.5 之后,code to video 到了新的高度。
复刻一个片段 → 完整复刻 10 分钟 → 创作新的 10 分钟。
@HyperFrames_ editor 已经在第三步了:「一口气了解泰国经济」,9 分 34 秒,一集她没讲过的题目。
最近 reverse engineering 很火,刷屏的小Lin说复刻帖我都看了。大多数复刻的是一段 1–2 分钟的 motion graphics:形似,但简单、廉价。小Lin的动效细节是满的,动作、编排、前后的连续性,全都和口播文案咬在一起。这不是复刻一分钟的片子能做出来的。
素材可以照着复刻。但如果从 0 开始呢?如果要一集接一集地做呢?
- 素材去哪找
- 每句话配什么画面,怎么卡着文案进出
- 文案怎么写才是小Lin味,而不是别的博主,或者流水线式的讲解
- 数字人和声音怎么撑满 10 分钟:语序、语调、停顿都要对
- 音效怎么提取、归类,用在对的地方
- 背景音乐在哪起、在哪停
到这一步,agent 就一头雾水了。
@HyperFrames_ editor 把这些都解决了:
- 把她拆成一套模板:怎么写、怎么剪、怎么动
- 动效逐帧对齐,1:1 重建
- 音效和音乐跟着口播卡点
- HeyGen 数字人 + 专业级声音克隆
下面泰国这集就是这么来的:同一套模板,换个题目就是新的一集。
HyperFrames editor 做的不是复刻小Lin说,是一个能无限往下做的「一口气了解 XXX」系列。hyperframes.dev/studio
(视频中的主持人形象和声音均为 AI 生成,非小Lin本人,仅作技术演示)
Miguel Ángel retweeted
Btw this video was created completely by head of marketing using our new @HyperFrames_ slack agent for teams
Let me know if you're interested in getting onboarded for your company!
Miguel Ángel retweeted
ChatGPT can drive Suno and HeyGen from the same chat now.
So obviously the first thing we did was invent a DJ and book him a main stage.
Try it for yourself and tag us.
Miguel Ángel retweeted
made a HyperFrames version of @stackedlol's ascii sharks 🦈
every shark is now a school of tiny Frameys
also, this video can be completely reversed and remixed in @HyperFrames_ studio
Miguel Ángel retweeted
Replying to @WaterrrMiao @HyperFrames_
好多人反馈的口型一致问题,是因为第一版编辑的时候agent改了两处声音,但没有重新对齐口型
现在我们用一个prompt最终配音从头在 HeyGen 重新生成了整段数字人,再由 agent 自动替换进片子里所有出镜的位置,现在是完整版本!!!
Miguel Ángel retweeted
HeyGen have been shipping like crazy.
Here is a round-up of their latest releases:
Hyperframes Studio
HeyGen Video
HeyGen Professional Voice
This entire video was edited with HeyGen products!
Miguel Ángel retweeted
No @HyperFrames_ launch today but don't worry we've got plenty coming this week
What do you all want to see us launch in @HyperFrames_ Studio??
Miguel Ángel retweeted
复刻小Lin说 1 分钟,现在谁都能做。
做满 10 分钟、还是那个味道,是另一回事🤥
Opus 5.5 之后,code to video 到了新的高度。
复刻一个片段 → 完整复刻 10 分钟 → 创作新的 10 分钟。
@HyperFrames_ editor 已经在第三步了:「一口气了解泰国经济」,9 分 34 秒,一集她没讲过的题目。
最近 reverse engineering 很火,刷屏的小Lin说复刻帖我都看了。大多数复刻的是一段 1–2 分钟的 motion graphics:形似,但简单、廉价。小Lin的动效细节是满的,动作、编排、前后的连续性,全都和口播文案咬在一起。这不是复刻一分钟的片子能做出来的。
素材可以照着复刻。但如果从 0 开始呢?如果要一集接一集地做呢?
- 素材去哪找
- 每句话配什么画面,怎么卡着文案进出
- 文案怎么写才是小Lin味,而不是别的博主,或者流水线式的讲解
- 数字人和声音怎么撑满 10 分钟:语序、语调、停顿都要对
- 音效怎么提取、归类,用在对的地方
- 背景音乐在哪起、在哪停
到这一步,agent 就一头雾水了。
@HyperFrames_ editor 把这些都解决了:
- 把她拆成一套模板:怎么写、怎么剪、怎么动
- 动效逐帧对齐,1:1 重建
- 音效和音乐跟着口播卡点
- HeyGen 数字人 + 专业级声音克隆
下面泰国这集就是这么来的:同一套模板,换个题目就是新的一集。
HyperFrames editor 做的不是复刻小Lin说,是一个能无限往下做的「一口气了解 XXX」系列。hyperframes.dev/studio
(视频中的主持人形象和声音均为 AI 生成,非小Lin本人,仅作技术演示)
if you are told to do "founder-led marketing" and have no idea how
use hyperframes dot studio
here's a "case study" of @jasonzhou1993 and how he used @HyperFrames_ to rack millions of views for @treg_ai























