Shadowing Practice: The mystery is solved... and the answer is 40x cheaper than Claude - Learn English Speaking with Video
Choose a practice mode
Each mode opens a full-screen practice room.
58 sentences
If sentences are too short or too long, click Edit to adjust them.
1
A few weeks ago, while I was busy putting the final touches on my Dolly Parton is too fast for the Grim Reaper tattoo, someone anonymously dropped a new Frontier-grade model online.
2
They called it Ox Alpha, and in its first six days it served 42 trillion tokens, which easily made it the most popular model on Open Router,
3
doubling the next closest, which was DeepSeek.
4
It comes with a million token context window, accepts video input, writes code like it has trouble making eye contact, and until just a few days ago, it was completely free.
5
Some assumed it was Zipu, others Xiaomi, and at least one guy was convinced that Google actually shipped something useful.
6
In today's video, we'll look at how this anonymous model won over the internet, who's behind it, and see if it can finally help me realize my horse SaaS dream.
7
It is September 1st, 2026, and you're watching the Code Report.
8
It hasn't even been two weeks since we talked about how
9
Silicon Valley is terrified of what's happening in China after the release of DeepSea Karnas.
10
And unfortunately, Dario and Sam's terrible, horrible, no-good, very-bad month may have gotten even worse because this new anonymous model had Made in China written all over it.
11
And as we know, things
12
that escape out of Chinese labs tend to have a big impact on the rest of the world.
13
The first clue was that AuxAlpha was the fifth anonymous animal model to show up on OpenRouter recently, and the four before it all turned out to be from China.
14
And even more convincing, some devs noticed that its stack traces and error codes matched Zipu's API, and a tokenizer fingerprint test came back matching their GLM series,
15
but nobody waited for a confirmation to start using it.
16
Yes, the fine print did say that all your prompts would be retained by a mysterious, probably foreign entity, but nobody actually cares about privacy when the magic box is free and works this well.
17
So devs happily pasted their company's proprietary code into it, routed cloud code through it to dodge their rate limits,
18
and at its peak, Aux Alpha accounted for nearly a third of Open Router's entire weekly traffic.
19
The hype machine got so out of control
20
that a misleading screenshot showing an 80% score on the Deep SWE benchmark went viral.
21
Even though the real number turned out to be closer to 58, and someone even speculated it was so good it might be a stolen Claude checkpoint.
22
Sure enough, on August 26th, Zipu came clean and announced that Aux Alpha was GLM 5.3 Flash,
23
a natively multimodal mixture of experts model with 320 billion total parameters, and its MIT-licensed weights hit hugging face the same day.
24
But the detail that really killed the vibe in Silicon Valley was
25
that the entire stealth run was served on just 100,000 Chinese-made chips, which I'm told really hurt Jensen's feelings.
26
Of course, the reveal also meant the free samples were over, and the anonymous endpoint was replaced with an actual pricing page,
27
but it's still stupid cheap at just 15 cents per million input tokens and 50 cents per million out.
28
And with a 50% discount through September 9th, it's currently up to 40 times cheaper than Claude.
29
And as we know, normal launches like this come with a bunch of trust me bro benchmarks
30
that usually fall apart the second anyone touches them, but their benchmarks hold up pretty well under outside scrutiny.
31
The catch is that it's slow, loves to yap, and occasionally gets stuck in doom loops, but in my opinion, there's only one way to test if it's actually any good,
32
and that's by using it on a deer project of mine.
33
After failing to raise funding for Horsetinder, I've pivoted to its once beloved cousin, Horsetube, a 2016 AngularJS app whose last commit was 10 years ago.
34
So I asked it to modernize the legacy stack without losing the humor and vibes I had so carefully crafted.
35
It correctly detected the outdated technologies
36
and then proceeded to narrate its plan of attack in every little detail for at least 20 minutes.
37
It then started building a design system in Raw Dog CSS.
38
It went with a beautiful purple slot background and rainbow gradient buttons, showing true inspiration, sorry, distillation from other popular LLM's visual tastes.
39
Because of its vision capability, it was able to detect a CSS overflow bug on mobile screens,
40
properly diagnose it, and fix it with the cursed min-width-equals-zero trick that always seemed to work.
41
It added some nice enter animations for video cards, which was a really nice touch.
42
All the bits of humor are still there, and it even added some new ones, like these toast notifications.
43
Code-wise, it replaced the legacy stack with no stack.
44
It wrote the app in HTML, CSS and vanilla JavaScript and wrote code that looks like a legacy code base of its own.
45
But I really wanted to test its vision skills, so I fed it a random horse video and asked it to add it to the app and create a title, description, and comments that are relevant to what's in the video.
46
It decided to use FFmpeg to break down the video into a series of frames, analyze those frames, and it did a really good job.
47
It got the full context of what happens in the video, nailed the description, and even aligned the humor with the rest of the platform.
48
And another tool checking out is Exa, the sponsor of today's video.
49
It's a search engine built for AI agents that gives you a lot more than just web results.
50
It includes 350 million research papers, financial data, and info on over 1 billion people and 70 million companies, with new sources added every week.
51
I hooked it up to PolyShip to let my users gamble on the stock market in even more degenerate ways, like betting on which tech CEOs will sell company stock right before their next earnings call,
52
and Exo was able to pull the recent SEC filings and convert it into structured JSON in one shot.
53
It also cites all of its sources, so users can't accuse me of cheating anymore.
54
You can plug directly into Cloud, ChatGPT, or your other favorite agent using their native plugins,
55
or try ExaAgent to get answers to more complex queries that need to pull from multiple datasets simultaneously,
56
which is a lot more powerful than what LLMs can do on their own.
57
Try Exa today for free and get $20 in credits at the link below.
58
This has been The Code Report, thanks for watching, and I will see you in the next one.
Download App
Everything you need to speak fluently
AI PronunciationScore every sentence
IPA PracticeMaster every sound
VocabularyBuild your word bank
Vocab GameLearn while playing
Same Channel
C12
The worst week in cybersecurity history... so far
Daily ConversationUniversity StudentsWorking Adults
C11
PostgreSQL in 100 Seconds
Daily ConversationWorking Adults
C11
React Native in 100 Seconds
Daily ConversationUniversity Students
C13
Every app on Earth is now open source... whether it likes it or not
Daily ConversationWorking Adults
C14
Brainf**k in 100 Seconds
Daily ConversationUniversity Students
C13
OpenAI’s lead engineer was secretly a spy...
Daily ConversationUniversity StudentsWorking Adults
C12
Mind-bending new programming language for GPUs just dropped...
Daily ConversationUniversity Students
C12
China just dropped a new frontier model... and nobody can explain how
Daily ConversationUniversity StudentsWorking Adults
C12
The last real programmers are dying out...
Daily ConversationUniversity StudentsWorking Adults
C11
Did the AI bubble just pop?
Daily ConversationUniversity StudentsWorking Adults
C130
Docker in 100 Seconds
American AccentBusiness EnglishUniversity StudentsWorking Adults
C16
The plan to break apart Google... RIP Chrome
American AccentBusiness EnglishDaily ConversationUniversity Students
B23
OpenAI’s new browser feels familiar…
American AccentBusiness EnglishUniversity StudentsWorking Adults
C19
The most interesting "hack" in history...
American AccentBusiness EnglishUniversity StudentsWorking Adults
C15
Tragic mistake... Anthropic leaks Claude’s source code
American AccentBusiness EnglishUniversity StudentsWorking Adults
C210
Angular in 100 Seconds
American AccentBusiness EnglishUniversity StudentsWorking Adults
B23
Firefox and Tor hit with 9.8 critical level exploit
American AccentBusiness EnglishUniversity StudentsWorking Adults
B23
5 open source tools that replaced my $320/mo AI stack...
American AccentBusiness EnglishUniversity StudentsWorking Adults
C19
When open-sourcing your code goes wrong...
American AccentBusiness EnglishUniversity StudentsWorking Adults
C12
C++ in 100 Seconds
American AccentBusiness EnglishUniversity StudentsWorking Adults
FOR YOU
Recommend Video
C11
764 sentences
The Youngest Self-Made Female Billionaire Behind Kalshi
Business EnglishUniversity StudentsWorking Adults
C13
57 sentences
Meta's new model wants "deep access" to your personal life...
American AccentBusiness EnglishUniversity StudentsWorking Adults
C1405 sentences
Supply Chain Expert Answers Chinese Manufacturing Questions | Tech Support | WIRED
Business EnglishUniversity StudentsWorking Adults
C11
462 sentences
Expensive RTX 5090 for LLMs? NO. Use This Instead. (Tesla V100 SXM2 + Z8 G4, #RACERRRZ)
Daily ConversationUniversity StudentsWorking Adults
C14
76 sentences
The Limitless AI Lie. The Bubble Is Slowly BURSTING.
Business EnglishDaily ConversationUniversity Students
C13
180 sentences
New FREE DeepSeek AI Coder is WILD!
Business EnglishUniversity StudentsWorking Adults
C11
155 sentences
Inside a scam call center
Business EnglishUniversity StudentsWorking Adults
C12
145 sentences
Most devs don't understand how LLM tokens work
British AccentBusiness EnglishUniversity StudentsWorking Adults
B22
243 sentences
I might have found the BEST budget router...
Business EnglishUniversity StudentsWorking Adults
C14
413 sentences
Exposing The Flaw In Our Phone System
Business EnglishUniversity StudentsWorking Adults
B21
190 sentences
This 25-Year-Old Built a 2-Year Moat Nobody Can Bet Against | Corgi, Nico Laqua & Emily Yuan
Business EnglishUniversity StudentsWorking Adults
Vocabulary and speaking notes for this lesson
This video has 58 sentences and 1121 words to shadow. The speech runs for 5:31. The speaker talks fast, about 203 words per minute, so expect linked and reduced sounds. Only 77% of the words are among the 3,000 most common in English, so the vocabulary is demanding.
Key vocabulary in this video
15 less common words from the video, with pronunciation and meaning:
- anonymous /əˈnɒn.ə.məs/ (adjective) — Lacking a name; not named, for example an animal not assigned to any species .
- stack /stæk/ (noun) — A large pile of hay, grain, straw, or the like, larger at the bottom than the top, sometimes covered with thatch.
- legacy /ˈlɛɡəsi/ (noun) — Money or property bequeathed to someone in a will.
- humour /ˈhjuːmɚ/ (noun) — The quality of being amusing, comical, funny.
- token /ˈtoʊkən/ (noun) — Something serving as an expression of something else.
- benchmark /ˈbɛn(t)ʃmɑːk/ (noun) — A standard by which something is evaluated or measured.
- cloud /ˈklaʊ̯d/ (noun) — A visible mass of water droplets suspended in the air.
- input /ˈɪnpʊt/ (noun) — The act or process of putting in; infusion.
- convince /kənˈvɪns/ (verb) — To make someone believe, or feel sure about something, especially by using logic, argument or evidence.
- detect /dɪˈtɛkt/ (verb) — To discover or find by careful search, examination, or probing.
- vibe /vaɪb/ (noun) — An atmosphere or aura felt to belong to a person, place or thing.
- dev (noun) — A monster, dragon, serpent, or giant.
- trick /ˈtɹɪk/ (noun) — Something designed to fool, dupe, outsmart, mislead or swindle.
- carefully /ˈkɛɹf(ə)li/ (adverb) — With care; attentively, circumspectly.
- horrible /ˈhɔɹ.ɪ.bəl/ (adjective) — Causing horror; terrible; shocking.
Phrasal verbs you will hear
- come with (verb) — To join and come along.
- break down (verb) — To stop functioning.
- check out (verb) — To record the departure or withdrawal of someone or something (such as guests, employees, books, etc.).
- come back (verb) — To return to a place.
- fall apart (verb) — To disintegrate, to break into pieces.
- get in (verb) — To get into or inside something, literally or figuratively.
- go with (verb) — To choose or accept (a suggestion).
- hold up (verb) — To wait or delay.
Pronunciation to watch
The speaker uses 6 contractions and reduced forms, such as can't, hasn't, I'm. Say them the short way, as you hear them.
- The “th” sounds: width /ˈwɪtθ/, stealth /stɛlθ/
- The “sh” and “zh” sounds: benchmark /ˈbɛn(t)ʃmɑːk/, inspiration /ˌɪn.spɪˈɹeɪ.ʃən/, occasionally /əˈkeɪʒənəli/, animation /ˌæn.ɪˈmeɪ.ʃən/, confirmation /ˌkɑn.fəɹˈmeɪ.ʃən/
- Long words — get the stress right: anonymous /əˈnɒn.ə.məs/, inspiration /ˌɪn.spɪˈɹeɪ.ʃən/, occasionally /əˈkeɪʒənəli/, animation /ˌæn.ɪˈmeɪ.ʃən/, mysterious /mɪˈstɪɹi.əs/
How to practise with this video
- Listen to the whole video once without speaking and note the words you do not know.
- Start at 0.75× speed, shadow it sentence by sentence, then go back to normal speed once it feels easy.
- Record yourself and compare with the original, paying attention to words like anonymous, stack, legacy.
Grammar in this video
The structures the speaker uses most, with the exact words from the video:
| Structure | In the video |
|---|---|
| Present perfect have/has + past participle — a past action that still matters now | hasn't even been · I've pivoted · has been |
| Passive voice be + past participle — the focus is on what happens, not who does it | would be retained · was served · was replaced |
| Relative clauses who / which + clause — extra information about a person or thing | closest, which was · cards, which was |
What is the Shadowing Technique?
Shadowing is a science-backed language learning technique originally developed for professional interpreter training and popularized by polyglot Dr. Alexander Arguelles. The method is simple but powerful: you listen to native English audio and immediately repeat it out loud — like a shadow following the speaker with just a 1–2 second delay. Unlike passive listening or grammar drills, shadowing forces your brain and mouth muscles to simultaneously process and reproduce real speech patterns. Research shows it significantly improves pronunciation accuracy, intonation, rhythm, connected speech, listening comprehension, and speaking fluency — making it one of the most effective methods for IELTS Speaking preparation and real-world English communication.