<p><pre><code> > we are open-sourcing two foundation models co-developed by Hy4 and Hyra 2.0: HyCell for single-cell representation and HyPDE for physical field modeling.
</code></pre>
What a refreshing contrast. Where American startups brag about stopping researchers from writing grants about neglected tropical diseases and making novel anti-depressants, these folks are open sourcing the means to do science. <a href="https://news.ycombinator.com/item?id=49651727">https://news.ycombinator.com/item?id=49651727</a><p>That's how you get a civilizational advantage. By allowing people to just do things. It's so bizarre to watch China embrace the wonders of science and technology, while certain American startups and communities repeat "WHAT ABOUT THE BIOWEAPONS?"<p>The American century has been fumbled hard.
> Looking ahead
> Although Hy4 is one of the strongest open-weight models available today, a notable gap remains between it and top proprietary models, particularly on challenging internal benchmarks for research-level mathematics, long-horizon coding, model self-improvement, and advanced visual question answering. We will continue to steadily scale pre-training and post-training to narrow this gap.<p>I like that they acknowledge that. For working with ai as a tool open models are already great. For millennium problems and doomermaxxing they are not there yet.
The game development focus is interesting:<p>> For game development, we partnered with professional developers at Tencent Interactive Entertainment Group (IEG) to curate high-quality training data and build evaluations. We focused on feature implementation, gameplay logic, state tracking, and asset pipelines. We also extensively adapted Hy4 for major engines such as Unreal Engine, Unity, and Godot, and expanded its capabilities to multiplayer game development. Compared with Hy4 preview, Hy4 improves by 19.1% on GameDev and 9.8% on GameLogic.<p>Here's more about that game development studio - <a href="https://en.wikipedia.org/wiki/Tencent_Games" rel="nofollow">https://en.wikipedia.org/wiki/Tencent_Games</a> - they've been around since 2004.
Tencent is one of the largest game development asset holders in the world. They're the Berkshire Hathaway of video games. They own, World of Warcraft (minority shareholder), League of Legends, Valorant, Warframe, Clash of Clans, Player Unknown Battlegrounds, FIFA Online, Pokemon Unite, Fortnite (minority shareholder) annd more. They also own 40% of Epic Games.
Landed on the cost/accuracy frontier in our data analytics benchmark: <a href="https://plotly.com/blog/tencent-hy4-plotly-data-analytics-bench/" rel="nofollow">https://plotly.com/blog/tencent-hy4-plotly-data-analytics-be...</a>, right in between GPT-6.1 Sol and GPT-6.1 Terra.<p>Verrrry thorough in its work and therefore verrrrry slow. It spent a long time with each question, running about 26 queries for each question, and in some questions it ran 64 queries. One of the queries that it wrote also took about an hour to run, which affected its overall exam time.<p>Last year the open weight models were the cheap & good alternatives. Then in September/October that cost story got rewritten by the GPT-6 series and Claude's 5 series.<p>Hy4 brings open models back on the cost/accuracy frontier, but its behavior is too thorough + slow.<p>Another thing we're seeing is that post training open models is starting to see results. Particle-1 from last week is also on our frontier, which is post-trained on DeepSeek V4.1 and Fireworks Ember-1 from a few weeks ago was post-trained on Kimi K3.<p>Take these numbers with a grain of salt: this exam is just for short horizon, low-context data analysis tasks, not full-fledged software projects.
Slightly worse than Kimi 3.1M Pro (which just released open weights today) but much cheaper even excluding the current 50% off promotion, and Tencent promises to make it even cheaper once it goes open-weight. Interesting price economics.
Refreshing to see normal looking bar charts for the benchmarks for once. Yes it shows they are in the middle of the pack but at least we can understand.
increasingly feeling like the entire US economy hangs on whether or not Ant/OAI can stay ahead of the Chinese models.<p>If the gap closes, we do zero austerity, and interest rates are high, I cannot see a way out of a major economic crisis, even if we do start to see dramatic productivity wins.<p>More on-topic (and maybe more positive?), but this model feels a bit more expensive than I would have expected!
> I cannot see a way out of a major economic crisis, even if we do start to see dramatic productivity wins.<p>The only thing worse for the economy than failure would be success.
i dont see the connection from 2 model providers relative position to the entire economy? especially in the face of dramatic productivity wins.
It's arguable that there is a threshold of quality that matter and US and China are both there already. I guess we'll see. Many of us on here are software engineers and I think we've seen that while Opus 4.7 (I think) was some sort of outlier/miracle, today there are several options to choose from. Depending on the task, the harness, etc Astra/Fable are only better when the work demands their level of capability - which much of software engineering does not.
Similar to Kimi K3.1 Pro, they nuked their cyber security skills too.
Similar to Kimi, they didn't "nuke" anything. They opted <i>not to train the model on cyber-offense tasks</i> which is a very reasonable choice when compared to what OpenAI and Anthropic do (train on cyber offense and then do a <i>surprised pikachu face</i> when their models go rouge and engage in cyber offense), and gives the model more scope for training on benign tasks.
The price numbers look weirdly specific, I wonder where they came from.<p>Personally, the thing I'm most excited about is the FrontierCode 2.0 score, which is a benchmark I trust a bit more - 31 seems to be roughly on par with 6.1 Sol and Opus 5.5 medium
Videos 1 and 4 on vision and computer use are wild. They depict the model finding a 2m feature from a large satellite image, then determining where a photo was taken. Is it just me or do the Chinese model releases have less concern about looking like a military weapon? I don't think it's a bad thing necessarily, it seems honest. Also the music they choose for it is hilarious.
Not cost competitive. The performance difference is not commensurate with the price.<p><pre><code> Model | Cached Input | Input (Miss) | Output
---------------|--------------|--------------|----------
Tencent Hy4 | $0.042 | $0.83 | $2.50
DeepSeek (peak)| $0.006 | $0.30 | $1.20
DeepSeek (off) | $0.003 | $0.15 | $0.60</code></pre>
We've moved past per token cost a long time ago. Smaller models use more tokens. What's interesting is the cost per task, on that front deepseek isn't particularly good.<p><a href="https://artificialanalysis.ai/#intelligence-efficiency-tabs" rel="nofollow">https://artificialanalysis.ai/#intelligence-efficiency-tabs</a>
It highly depends on the task. So far, on the tasks I personally benchmark, DS v4.1 Flash @ deepinfra.com is always cheaper because their cached input price for it is so very low and their cache hit rate is so very high.
Wild that this is the first question I ask at this point. After that it’s a question of how it stacks intelligence wise vs glm or ds.<p>OAI / Ant are in for a reckoning.
How does it do with tasks anthropic doesn’t like? This is my main use case for open models.
<i>>We filtered out offensive content such as exploit development, reverse engineering, and penetration testing to avoid amplifying potential offensive capabilities.</i><p>So, yeah. Back to GLM 5.3.
Another model, another broken website. I can't even zoom on the benchmarks.<p>Le cordonnier mal chaussé.
[dead]