Hugging Face has published download statistics for January–July 2026, revealing that the Chinese open model Qwen racked up nearly twice as many downloads as Gemma. The thread saw a split of opinions over the technical standing of Chinese-made AI, alongside a lively exchange of real-world stories about switching to local LLMs and tips for managing subscriptions.
Qwen, Kimi, MiniMax… the momentum behind Chinese AI models these days shows no sign of stopping. Statistics on open models released by Hugging Face on August 14 (US time) reveal that Chinese AI models are continuing to expand their share.
This report comes from Hugging Face, where a huge number of AI models are uploaded, and covers data from January 2026 through July 31. Among the notable trends observed during this period is the following.
The rise in frontier models
Source: pc.watch.impress.co.jp / Original article here
Guess it's about picking the right tool for the job.
For chat, Gemma's Japanese is better.
For coding, Qwen's the way to go.
For chat, Gemma's Japanese is better.
For coding, Qwen's the way to go.
>>2
That's exactly how it feels using them, no joke.
That's exactly how it feels using them, no joke.
>>1
Hold on now.
Qwen's been around way longer than Gemma — since April 2023, right from the start.
It's been the king of open models since early on, right after Meta's Llama in February 2023.
Gemma didn't show up until a year later, in April 2024.
If anything, the story should be about how much Gemma has caught up.
"The rise of China" this, "China's latest thing" that —
Impress Watch has gotten pretty dumb too, huh.
Hold on now.
Qwen's been around way longer than Gemma — since April 2023, right from the start.
It's been the king of open models since early on, right after Meta's Llama in February 2023.
Gemma didn't show up until a year later, in April 2024.
If anything, the story should be about how much Gemma has caught up.
"The rise of China" this, "China's latest thing" that —
Impress Watch has gotten pretty dumb too, huh.
>>21
For PC or programming questions, even a 9B model can hold a conversation at a level where you don't need the internet, but LLMs just can't handle current-events questions.
Ask a local LLM about the latest news and it treats it as fake and just refuses to answer at all.
>>4
Some "king" — the moment they went closed, they lost consumer interest instantly and had to scramble back to the open-source route.
For PC or programming questions, even a 9B model can hold a conversation at a level where you don't need the internet, but LLMs just can't handle current-events questions.
Ask a local LLM about the latest news and it treats it as fake and just refuses to answer at all.
>>4
Some "king" — the moment they went closed, they lost consumer interest instantly and had to scramble back to the open-source route.
My company has already fully switched over to Chinese AI. Great value, great performance — anyone who hasn't switched yet is hopelessly behind the times.
>>11
Must be nice doing work where it's fine if your data gets siphoned off.
Must be nice doing work where it's fine if your data gets siphoned off.
>>11
A publicly listed company couldn't possibly place an order with them.
A publicly listed company couldn't possibly place an order with them.
Once local LLMs get good enough, you won't need the cloud-based ones anymore.
Chinese-made AI only gives answers that align with what the Chinese Communist government wants. There's no way something like that can be trusted.
>>22
Who's using an LLM to ask that kind of question anyway lol
Who's using an LLM to ask that kind of question anyway lol
If you think of right now as the golden era of generosity, you'd better download and keep a copy of Gemma 4 too. Even pure science questions like "how is a thermos made" or "is the vacuum seal held with lead solder" will eventually get nothing but ad-sanitized answers.
>>29
Who do you think is the one serving Gemma in the first place?
You don't even understand what "open model" means,
and you're out here saying "you'd better…"
Who do you think is the one serving Gemma in the first place?
You don't even understand what "open model" means,
and you're out here saying "you'd better…"
I'm actually cancelling my Claude $200 plan at the end of this month.
Even on the $200 plan, Fable 5 would hit a week's worth of rate limits after running for just a single day — totally unusable.
So starting this month I switched to Kimi K3's $99 plan instead, and the performance is basically on par with Fable 5, except now I don't have to worry about rate limits.
And as of yesterday I've got Qwen 3.8 27B running locally, handing it lighter tasks like coding.
Qwen 3.8 27B performs about the same as Opus 4.8, so I've got nothing to complain about.
Honestly, I just want to say thank you, China.
Even on the $200 plan, Fable 5 would hit a week's worth of rate limits after running for just a single day — totally unusable.
So starting this month I switched to Kimi K3's $99 plan instead, and the performance is basically on par with Fable 5, except now I don't have to worry about rate limits.
And as of yesterday I've got Qwen 3.8 27B running locally, handing it lighter tasks like coding.
Qwen 3.8 27B performs about the same as Opus 4.8, so I've got nothing to complain about.
Honestly, I just want to say thank you, China.
>>30
Thanks for the report. Gonna copy what you're doing.
Thanks for the report. Gonna copy what you're doing.
>>30
Read this and tried it on my own machine (Ryzen AI Pro 370), but Qwen 3.8 27B only managed 6 tokens/s even at Q3_K_M lol
Memory bandwidth really is everything.
Guess I'll quietly go back to Ornith 1.0-9B.
Read this and tried it on my own machine (Ryzen AI Pro 370), but Qwen 3.8 27B only managed 6 tokens/s even at Q3_K_M lol
Memory bandwidth really is everything.
Guess I'll quietly go back to Ornith 1.0-9B.
Believing the hype, setting up a local model, and then finding out it just doesn't cut it — everyone goes through that phase, huh.
Claiming the same performance as a frontier model when the parameter count is nowhere close is a stretch to begin with.
Claiming the same performance as a frontier model when the parameter count is nowhere close is a stretch to begin with.
>>38
Parameter count is about the amount of knowledge, not a ceiling on reasoning ability like you'd need for coding.
Parameter count is about the amount of knowledge, not a ceiling on reasoning ability like you'd need for coding.
>>42
Saying reasoning doesn't come from knowledge — typical of Japan, land of the flash-quiz trivia shows. (a jab at Japan's love of quick-recall trivia/quiz TV culture)
Saying reasoning doesn't come from knowledge — typical of Japan, land of the flash-quiz trivia shows. (a jab at Japan's love of quick-recall trivia/quiz TV culture)
>>42
"Parameter count is the amount of knowledge" is completely wrong. Go read a textbook.
"Parameter count is the amount of knowledge" is completely wrong. Go read a textbook.
Burning through $200 in a single day — what are you even doing to use that much?
Like a lot of people have pointed out, no matter how cheap the pay-as-you-go token price gets, it's hard to beat subscription plans on cost, OpenAI included.
I'd get it if the issue were licensing restrictions, but…
There's also the question of what you do once the subscription's gone.
Like a lot of people have pointed out, no matter how cheap the pay-as-you-go token price gets, it's hard to beat subscription plans on cost, OpenAI included.
I'd get it if the issue were licensing restrictions, but…
There's also the question of what you do once the subscription's gone.
>>40
Uh, you do know that using the AI agent API through Claude Code is billed sep-a-rate-ly, right?
Uh, you do know that using the AI agent API through Claude Code is billed sep-a-rate-ly, right?
>> The rise of Chinese AI
There's no "rise" when it's just benchmarks with no real track record.
There's no "rise" when it's just benchmarks with no real track record.
It literally says right at the top that it's built from Gemma 4 and Qwen 3.5.
State-of-the-Art Coding Agents:
Available in 9B-Dense, 31B-Dense, 35B-MoE, and 397B-MoE
(post-trained on top of Gemma 4 and Qwen 3.5),
achieving state-of-the-art performance among open-source models of comparable size on coding benchmarks such as Terminal-Bench 2.1, SWE-Bench, NL2Repo and OpenClaw.
State-of-the-Art Coding Agents:
Available in 9B-Dense, 31B-Dense, 35B-MoE, and 397B-MoE
(post-trained on top of Gemma 4 and Qwen 3.5),
achieving state-of-the-art performance among open-source models of comparable size on coding benchmarks such as Terminal-Bench 2.1, SWE-Bench, NL2Repo and OpenClaw.
>>48
https://huggingface.co/ornith-ai/Ornith-1.0-9B
Just on Hugging Face alone this has 2.4 million downloads, so that's pretty impressive.
https://huggingface.co/ornith-ai/Ornith-1.0-9B
Just on Hugging Face alone this has 2.4 million downloads, so that's pretty impressive.
>>56
Even running an agent loop, it doesn't beat that number. How many sessions are you running?
Even running an agent loop, it doesn't beat that number. How many sessions are you running?
>>60
Wait, are you running it single-threaded?
What a waste — people are apparently running 100 or 200 in parallel nonstop.
Wait, are you running it single-threaded?
What a waste — people are apparently running 100 or 200 in parallel nonstop.
I'm curious how much speed you actually get depending on the model/GPU combo.
Would be a shame to build a dual-GPU rig and have it turn out underwhelming.
So I want to try it on an overseas GPU cloud, but I'm scared to enter my credit card info.
And apparently you need to charge up something like 10,000 yen (~$65) before payment even goes through, which is even scarier.
By the way, you can pick GeForce-series GPUs too, but using those in a data center is against the license terms (except for blockchain use), so any provider offering that is sketchy.
I thought Sakura Internet's GPU servers would be cheap and safe, but then there was that personal data leak thing…
So, guess I'll just rent an AWS GPU instance to test it out.
Would be a shame to build a dual-GPU rig and have it turn out underwhelming.
So I want to try it on an overseas GPU cloud, but I'm scared to enter my credit card info.
And apparently you need to charge up something like 10,000 yen (~$65) before payment even goes through, which is even scarier.
By the way, you can pick GeForce-series GPUs too, but using those in a data center is against the license terms (except for blockchain use), so any provider offering that is sketchy.
I thought Sakura Internet's GPU servers would be cheap and safe, but then there was that personal data leak thing…
So, guess I'll just rent an AWS GPU instance to test it out.
>>68
Quit yapping and just go do it already.
That level of worry was a 2023 thing.
You're three years behind too.
Quit yapping and just go do it already.
That level of worry was a 2023 thing.
You're three years behind too.
What's an "artificial intelligence system"? Is that just local LLMs?
>>72
In this context it just means LLMs. Adding "local" would be wrong.
In this context it just means LLMs. Adding "local" would be wrong.
Japan's getting left in the dust, huh — behind by generations of models?
Guess we're an AI backwater now, better go beg China for help lol
Guess we're an AI backwater now, better go beg China for help lol
>>74
Japan's being left behind?
Then which country do you think Gemma is even from?
I'd love to hear your answer.
Japan's being left behind?
Then which country do you think Gemma is even from?
I'd love to hear your answer.
*This article is composed as an excerpt and summary of the 5ch (Business News+) thread “[AI] Qwen Nearly Doubles Gemma — The Rise of Chinese AI as Seen in Hugging Face Statistics.”
Leave a Reply