Qwen3-Next
Hacker ApprovedRanked 41 of 72 language models by recommenders
Many people run it locally on laptops or GPUs and find it fast and capable, though critics say it falls short for agentic coding.
Many people run it locally on laptops or GPUs and find it fast and capable, though critics say it falls short for agentic coding.
19 people gave an opinion
- recommend it
- 14
- criticize it
- 5
- approval share
- 74%
Recommended in 9 of 13 months
- November 2025: not recommended
- December 2025: recommended
- January 2026: recommended
- February 2026: recommended
- March 2026: recommended
- April 2026: recommended
- May 2026: not recommended
- June 2026: not recommended
- July 2026: not recommended
- August 2026: recommended
- September 2026: recommended
- October 2026: recommended
- November 2026: recommended (month in progress)
Why it carries the badge
- 14 people recommend itNeeds 10 or more, met.
- 2.8 recommenders for every criticNeeds 2 to 1 or better, met.
- Recommended in 9 of 13 monthsNeeds 4 or more, met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Runs locally on laptops and GPUs
“the mixture of experts models have been nice when i can run the dense reasoning layers on the gpu”
Comment from 25 Aug 2026 in “Qwen 3.8-Flash-Next releasing tomorrow (125B a6B)” “It's quite capable in running or training smart LLMs like Qwen3-Next-80B-A3B-Instruct-NVFP4”
Comment from 1 Jan 2026 in “Dell's version of the DGX Spark fixes pain points” “thanks to its new architecture, is quite fast even if part of it is streaming from disk”
Comment from 18 Sep 2026 in “Bonsai 2 27B: Near-Lossless Compression in a 9x Smaller Footprint” - Comment from 11 Feb 2026 in “GLM-5: From Vibe Coding to Agentic Engineering”
- Comment from 10 Jan 2026 in “What Claude Code Sends to the Cloud”
- Comment from 8 Jan 2026 in “Chinese AI models have lagged the US frontier by 7 months on average since 2023”
Solid output quality for its size
“Qwen3-Next and Nemotron 3 Nano are solid players in their respective local weight classes”
Comment from 13 Jan 2026 in “Anthropic Made a Big Mistake” “I have no issues believing it would be fine for more than just pet projects.”
Comment from 1 Nov 2026 in “Memdoor: A team of 5 Qwen3.8-27B models beats Fable 5 on LiveCodeBench Hard” “they have been indiscernible in terms of output”
Comment from 23 Oct 2026 in “Kimi K3.1 scores 55 on Artificial Analysis, placed on the Pareto line” - Comment from 18 Sep 2026 in “Qwen 3.8 Omni Flash”
Fast responses, even when streaming
“thanks to its new architecture, is quite fast even if part of it is streaming from disk”
Comment from 18 Sep 2026 in “Bonsai 2 27B: Near-Lossless Compression in a 9x Smaller Footprint” “works decently in my testing, and fast”
Comment from 22 Dec 2025 in “GLM-4.7: Advancing the Coding Capability”
What people warn about
Each mark is one comment
Weak performance against current models
“It was nowhere close neither in the output quality nor in performance to any current models”
Comment from 28 Feb 2026 in “We Will Not Be Divided” “Qwen 3.8 Next did work, but the performance was abysmal.”
Comment from 11 Sep 2026 in “DeepSeek launching v4.1 flash cheaper and more capable than v4 pro” Ignores instructions and debugs unprompted
“it has a realization that it would be simpler to just do the thing I told it not to do”
Comment from 4 Mar 2026 in “Something is afoot in the land of Qwen” “they both try to debug bugs they encounter unprompted”
Comment from 1 Nov 2026 in “GPT-5.6 Sol can beat Factorio on peaceful mode”
Other language models people recommend
The four most recommended other language models, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| Claude Opus | 3,331 | 2,746 | 55% | 13 | 4 mo156 owners | — |
| OpenAI GPT | 1,205 | 987 | 55% | 13 | 6 mo47 owners | — |
| Claude Sonnet | 854 | 586 | 59% | 13 | 6 mo46 owners | — |
| DeepSeek V4 Flash | 828 | 223 | 79% | 8 | 1 mo24 owners | Approved |
| Qwen3-NextThis page | 14 | 5 | 74% | 9 | — | Approved |
What people said
20 of the 23 comments that mention it, newest first. Each links to the original comment.
“I have no issues believing it would be fine for more than just pet projects.”
“they both try to debug bugs they encounter unprompted”
“they have been indiscernible in terms of output”
“thanks to its new architecture, is quite fast even if part of it is streaming from disk”
“No, it is actually very good.”
“Qwen 3.8 Next did work, but the performance was abysmal.”
“the mixture of experts models have been nice when i can run the dense reasoning layers on the gpu”
“it has a realization that it would be simpler to just do the thing I told it not to do”
“Qwen3-next, Qwen3.5, and Nemotron 3 Nano are all great examples.”
“It was nowhere close neither in the output quality nor in performance to any current models”
“it unfortunately doesn't have a vision encoder”
“the lastest 80B Qwen Next is not far from Opus 4.6 but runs on my laptop.”
“Qwen3-Next and Nemotron 3 Nano are solid players in their respective local weight classes”
“My personal LLM preference is for Qwen3-Next-80B with 4bit quantization”
“very, very useful if you have no internet”
“It's quite capable in running or training smart LLMs like Qwen3-Next-80B-A3B-Instruct-NVFP4”
“although it is not good enough for agentic coding”
“It may run a lot faster, though I don't know how well it does coding tasks.”
“works decently in my testing, and fast”
“Qwen3-Next was able to reuse the same prompts I used for Gemma3 perfectly”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.