Qwen3-30B-A3B
Hacker ApprovedRanked 24 of 72 language models by recommenders
Fast mixture-of-experts model that runs on modest hardware and handles everyday text and quick coding well; critics say it is verbose and weak on fiction and harder code.
Fast mixture-of-experts model that runs on modest hardware and handles everyday text and quick coding well; critics say it is verbose and weak on fiction and harder code.
45 people gave an opinion
- recommend it
- 41
- criticize it
- 4
- approval share
- 91%
Recommended in 10 of 13 months
- November 2025: recommended
- December 2025: recommended
- January 2026: recommended
- February 2026: recommended
- March 2026: recommended
- April 2026: recommended
- May 2026: recommended
- June 2026: recommended
- July 2026: recommended
- August 2026: recommended
- September 2026: not recommended
- October 2026: not recommended
- November 2026: not recommended (month in progress)
Why it carries the badge
- 41 people recommend itNeeds 10 or more, met.
- 10 recommenders for every criticNeeds 2 to 1 or better, met.
- Recommended in 10 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
Fast token generation on modest hardware
“It's very very fast and pretty good, and speed matters when you're replacing basic google searches and text manipulation.”
Comment from 21 Nov 2025 in “Olmo 3: Charting a path through the model flow to lead open-source AI” “Qwen models are in general amazing, but 30B is v fast cuz it's an MoE”
Comment from 21 Nov 2025 in “Olmo 3: Charting a path through the model flow to lead open-source AI” “I get a highly usable 50 tokens per second out of the A3B instruct”
Comment from 1 Feb 2026 in “Ask HN: Any real OpenClaw (Clawd Bot/Molt Bot) users? What's your experience?” - Comment from 6 Jan 2026 in “A 30B Qwen Model Walks into a Raspberry Pi and Runs in Real Time”
- Comment from 11 Nov 2025 in “Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch”
- Comment from 4 Feb 2026 in “Claude Code: connect to a local model when your quota runs out”
Good enough for everyday tasks
“I find the output of Qwen3's 30B model for that sort of task is plenty good enough.”
Comment from 22 Dec 2025 in “AI's real superpower: consuming, not creating” “it still offers me value for my use cases (many small prototypes and tests)”
Comment from 8 Dec 2025 in “The "confident idiot" problem: Why AI needs hard rules, not vibe checks” “Qwen3-30B which is good enough for some quick javascript or Python”
Comment from 5 Feb 2026 in “Claude Code: connect to a local model when your quota runs out” - Comment from 23 May 2026 in “Microsoft Drops Claude Code After Budget Overrun”
Runs on affordable consumer GPUs
“Qwen3 8B or A30B are quite capable models which run on a lot of hardware even today”
Comment from 23 Dec 2025 in “Local AI is driving the biggest change in laptops in decades” “even if the model doesn't completely fit on the GPU it's still fast enough”
Comment from 28 Feb 2026 in “How do I cancel my ChatGPT subscription?” “That model only needs a £500 second hand GPU - that's impressive for me.”
Comment from 5 Dec 2025 in “We gave 5 LLMs $100K to trade stocks for 8 months” - Comment from 15 Nov 2025 in “HipKittens: Fast and furious AMD kernels”
Useful for coding with prepared prompts
“the most efficient way is to use larger models for coordination, but a heretic'ed qwen 30B model for the implementations”
Comment from 21 Jul 2026 in “Agent swarms and the new model economics” “very useful for coding if I prepare a one-shot prompt with directions and relevant code”
Comment from 23 Dec 2025 in “GLM-4.7: Advancing the Coding Capability” “Qwen3-30B which is good enough for some quick javascript or Python”
Comment from 5 Feb 2026 in “Claude Code: connect to a local model when your quota runs out”
What people warn about
Each mark is one comment
Weak for coding and fiction
“didn't find it to be great for my coding use case”
Comment from 8 Jan 2026 in “A 30B Qwen Model Walks into a Raspberry Pi and Runs in Real Time” “absolutely terrible at editing fiction”
Comment from 21 Feb 2026 in “Ggml.ai joins Hugging Face to ensure the long-term progress of Local AI”
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 |
| Qwen3-30B-A3BThis page | 41 | 4 | 91% | 10 | — | Approved |
What people said
40 of the 54 comments that mention it, newest first. Each links to the original comment.
“the most efficient way is to use larger models for coordination, but a heretic'ed qwen 30B model for the implementations”
“given that they can already run Qwen3 30B A3B”
“Running Qwen3-30B-A3B-Instruct-2507-AWQ-4bit on an Olares One”
“are enough for pretty much all tasks I threw at them”
“you are missing some open source models (qwen 30b 3a, Deepseek 4 flash)”
“which offer a good balance of prompt response speed and output quality”
“Specific models & sizes for specific use cases on specific hardware at specific speeds.”
“the previous model is pretty fast but kinda dumb”
“they claim 3-6x inference thorough-put compared to Quen3-30B-A3B”
“only 30b a3b is smart enough not miss job detail (most of time)”
“even if the model doesn't completely fit on the GPU it's still fast enough”
“absolutely terrible at editing fiction”
“Qwen 30B and GLM Flash (also around 30B) are both very good for example and I use them regularly.”
“Qwen3-30B which is good enough for some quick javascript or Python”
“Qwen3 variants perform very well, speed wise.”
“Even Qwen3-30B-A3B performs better than Gemini, in my experience.”
“Best to run a smaller model like qwen3-32b/30b at Q6”
“That's my main complaint about qwen3:30b, how verbose its reasoning is.”
“I get a highly usable 50 tokens per second out of the A3B instruct”
“didn't find it to be great for my coding use case”
“Seems quite good, actually—much more useful than llama 3.2:3b, which has comparable performance on this Pi.”
“the biggest models you can run at a reasonable price right now are quantized versions of things like the Qwen3 30B A3B family”
“While still impressive nevertheless, being able to get 8T/s for a "A3B" compared to a dense 30B is very different.”
“6-7 tokens/sec is quite remarkable for the hardware.”
“Seems quite good, actually—much more useful than llama 3.2:3b”
“It does take up a little over 10 GB of RAM (watching with btop) before it segfaults and quits.”
“very useful for coding if I prepare a one-shot prompt with directions and relevant code”
“Qwen3 8B or A30B are quite capable models which run on a lot of hardware even today”
“I find the output of Qwen3's 30B model for that sort of task is plenty good enough.”
“Why would you use an Omni model for text only workload... There is Qwen3-30B-A3B.”
“it still offers me value for my use cases (many small prototypes and tests)”
“That model only needs a £500 second hand GPU - that's impressive for me.”
“any low quant low size model you're going to run locally”
“Qwen models are in general amazing, but 30B is v fast cuz it's an MoE”
“It's very very fast and pretty good, and speed matters when you're replacing basic google searches and text manipulation.”
“maintained 96.9% accuracy with graceful degradation”
“it's a 30b param model that fits into an AMD card and runs great”
“does ~27 tokens/sec on pure CPU on a $450 mini PC”
“Qwen 30B A3B is the only one that cutely starts on the landscaping and background.”
“It feels on par with the premium models, tbh.”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.