Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency. Summaries are generated from a stored copy of the article text; every sentence cites a span of that text, and nothing is shown that failed the citation check.
Official2
qwen.ai2026-02-16officialauthorityHN front pageabout the Qwen3.5 familyHN 434 pts · 214 comments17,678 words
Qwen3.5-397B-A17B is the first open-weight native vision-language model in the Qwen3.5 series, released on 2026/02/15. Built on a hybrid architecture fusing linear attention via Gated Delta Networks with a sparse mixture-of-experts, it activates just 17 billion parameters per forward pass despite comprising 397 billion total parameters. The model expanded language support from 119 to 201 languages while matching the >1T-parameter Qwen3-Max-Base capability.written, reviewed, cited
- The hosted Qwen3.5-Plus model defaults to a 1M context window and includes official built-in tools with adaptive tool use.
- For the BrowseComp benchmark, the model achieved a score of 69.0 using simple context-folding, while the discard-all strategy reached 78.6.
unsloth.ai2026-03-04authorityHN front pageabout the Qwen3.5 familyHN 416 pts · 107 comments729 words
Engineers can fine-tune the Qwen3.5 model family (0.8B, 2B, 4B, 9B, 27B, 35B-A3B, 122B-A10B) with Unsloth, which trains models 1.5 times faster using 50% less VRAM than FA2 setups. VRAM requirements for bf16 LoRA range from 3GB for the 0.8B model to 56GB for the 27B model, while the 35B-A3B variant operates on 74GB VRAM and the 122B-A10B on 256GB VRAM. MoE fine-tuning leverages a recent update delivering over 35% less VRAM usage and approximately 6x longer context support.written, reviewed, cited
- To preserve reasoning ability, mix reasoning-style examples with direct answers while keeping a minimum of 75% reasoning.
- Router-layer fine-tuning is disabled by default for stability on MoE models like Qwen3.5‑35B‑A3B.
- You can run Qwen3.5 RL with Unsloth even though it is not supported by vLLM by setting fast_inference=False.
Analysis6
github.com2026-09-21authorityHN front pageabout the Qwen3.5 familyHN 463 pts · 210 comments6,746 words
Kev is a family of small decision models built on Qwen3.5 and Qwen3.8 that you can train and run locally. The API matches TypeSafe's System One, so you can point their Python SDK at your local server. Models use a small adapter on a frozen base, except Kev-27B which ships as 51 GB of full weights. The 27B model achieves 0.851 accuracy on new sources.written, reviewed, cited
- The Kev-27B model requires 96–128 GB of RAM on Macs, a requirement marked as expected but not measured.
- An explicitly requested accelerator fails instead of falling back if the probe times out or crashes.
github.com2026-04-20authorityHN front pageHN 165 pts · 52 comments2,034 words
Lucebox is a speculative inference server for heterogeneous machines and consumer GPUs featuring custom kernels for speculative prefill and decoding. It delivers 208.1 tok/s average on Qwen 3.8 27B on one R9700 and achieves 6.1× prefill speedup for Laguna XS 2.1 33B at 256K on an RTX 3090. The system includes PFlash, a lossy prompt compression tool that should be disabled for exact-retrieval benchmarks.written, reviewed, cited
- The engine selects a discrete GPU over an integrated one before choosing the largest available device.
- The server automatically sizes the shared KV pool based on the available GPU memory.
github.com2026-10-04authorityabout the Qwen3.5 familyHN 2 pts · 0 comments1,558 words
llm.vhdl is a full-fabric VHDL inference engine for Qwen3.5-class models on Xilinx XCVU33P and XCVU35P FPGAs using INT4 streaming matvecs and HBM-resident weights. It supports single-card 9B or two-card 27B configurations, achieving about 2.5 tokens/s on silicon at 75 MHz. The 27B target fits a VU35P at about 46% LUT but requires the Jungle Cat board, which lacks PCIe and relies on slow Ethernet or JTAG for weight loading.written, reviewed, cited
- Reconfiguring the FPGA tears the PCIe bus down and rescans it, which can hang the host if a transaction faults.
- The inter-die link for the two-die residual hand-off cannot come up because the carrier's refclk oscillator sites are unpopulated.
- The Jungle Cat board lacks PCIe, forcing weight loading over a CoE/JTAG bridge that operates at roughly 1 MB/s.
sharpai.org2026-03-20HN front pageabout the Qwen3.5 familyHN 173 pts · 152 comments1,458 words
gslaller.github.io2026-09-29about the Qwen3.5 familyHN 2 pts · 0 comments
unsloth.ai2026-02-28authorityabout the Qwen3.5 familyHN 5 pts · 0 comments
Community2
news.ycombinator.com2026-08-06about the Qwen3.5 familyHN 4 pts · 2 comments157 words
This American (San Francisco) model, Laguna S 2.1:118B-a9B, outperforms Qwen3.5:122B-a10B and Sonnet in current benchmarking tests. It demonstrates better personality and natural creativity when designing UIs. Benchmarks need to include a minimum of hardware requirements, which are currently missing.written, reviewed, cited
news.ycombinator.com2026-04-16about the Qwen3.5 familyHN 7 pts · 2 comments
