AI tool comparison
Darwin-4B-David vs Tencent Hy3-preview
Which one should you ship with? Here is the side-by-side panel verdict, pricing read, reviewer split, and community vote comparison.
AI Models
Darwin-4B-David
4.5B merged model beats Gemma-4-31B on GPQA — no training needed
75%
Panel ship
—
Community
Paid
Entry
Darwin-4B-David is a 4.5-billion-parameter model that achieves 85.0% on GPQA Diamond — outperforming Google's Gemma-4-31B (84.3%) at roughly 1/7th the parameter count. The kicker: it required no training whatsoever. It was built in 45 minutes on a single H100 using MRI-guided DARE-TIES model merging, a novel variant of the merge-and-trim technique. The MRI-guided approach uses activation analysis to identify which parameters in each source model are most critical, then applies DARE-TIES merging only to the high-value weight regions. This avoids the catastrophic interference that usually degrades merged models. The result is a small model that inherits the strengths of multiple larger predecessors without any of the compute cost of fine-tuning. For the AI community, this is a meaningful data point: model merging continues to close the gap with expensive training runs. Darwin-4B-David demonstrates that thoughtful merge strategies can extract benchmark-level performance from models that are a fraction of the size, making capable AI more accessible on consumer hardware.
AI Models
Tencent Hy3-preview
Tencent's first open-source frontier MoE — 295B params, 21B active, free on HuggingFace
75%
Panel ship
—
Community
Free
Entry
Tencent's Hy3-preview is the company's first public frontier-class language model, released April 23 as open weights on Hugging Face. The model is a 295B parameter Mixture-of-Experts architecture with only 21B parameters active per token — keeping inference costs comparable to much smaller dense models while reaching capabilities that compete with leading proprietary systems. The release comes under new leadership: Yao Shunyu, a former OpenAI researcher, joined Tencent in early 2026 to build out its frontier AI effort. The team claims to have gone from project start to public release in under three months — an unusually fast timeline for a model of this scale. The 256K context window and strong performance on agentic and coding benchmarks position it directly against GLM-5.1 and Qwen3.6 in the open-source frontier race. Free inference is available on OpenRouter's free tier at launch, with the model also appearing on Hugging Face's Inference API. The architecture uses 192 routed experts in a hybrid dense-MoE configuration. For teams needing a capable open-weights model for agentic workflows without paying proprietary API rates, Hy3-preview arrives as a credible option at a remarkable cost-to-capability ratio.
Reviewer scorecard
“45 minutes on a single H100 to beat a 31B parameter model? That's an extraordinary efficiency ratio. MRI-guided merging is a technique I'll be watching closely. If this holds up across more benchmarks, it fundamentally changes how teams should think about building capable small models.”
“295B MoE with 21B active per token is a sweet spot for production use — you get frontier-quality outputs at a fraction of the compute cost. The 256K context and agent-optimized design make this immediately useful for complex workflow automation. Worth running evals against your specific use case.”
“GPQA Diamond is one benchmark. One. Benchmark performance doesn't translate linearly to real-world task performance, especially for a merged model that hasn't been fine-tuned for instruction following or RLHF alignment. Impressive number, but I'd want to see this on coding, reasoning chains, and RAG tasks before getting excited.”
“Tencent hasn't published a full technical report yet, so benchmark claims are hard to independently verify. The 'three months to frontier' narrative sounds impressive but raises questions about training data sourcing and evaluation rigor. Preview releases from large Chinese labs have historically required patience before production stability.”
“Model merging is the dark horse of AI efficiency research. If MRI-guided DARE-TIES merging can reliably produce results like this, it suggests we're nowhere near the ceiling for extracting value from existing open-weight models. The future may involve less training and more intelligent composition.”
“The pace of open-source frontier models from Chinese labs is accelerating faster than anyone predicted — we now have credible open-weight competition from Alibaba, Zhipu, Xiaomi, and Tencent simultaneously. This is geopolitically significant and means the open-source ecosystem will stay competitive with proprietary models for years.”
“A capable model in the 4-5B range that can run on a MacBook M-series is exactly what solo creators need for on-device inference. If Darwin-4B-David's performance holds on creative tasks, it's a genuine local creative AI for people without cloud budgets.”
“For multilingual creative work — especially for Chinese market content — having a frontier-quality open-source model from a Chinese lab is meaningful. The free OpenRouter tier means creators can experiment without API budgets.”
Weekly AI Tool Verdicts
Get the next comparison in your inbox
New AI tools ship daily. We compare them before you waste an afternoon.