LOADING DATASHEET
GROUNDED IN REAL HUMAN CONVERSATIONSYOUR BENCH
Qwen3.6 35B A3B
◆ OPEN
by Alibaba
OPEN WEIGHTSLOCALAPILONG CONTEXTPAIDREASONING
RANKED #34 OF 267 ASSISTANTS · OVERALL #51 OF 6,566 · VIBE SCORE 6.8 · 550 VOICES
Qwen3.6-35B-A3B is an open-weight multimodal model from Alibaba Cloud with 35 billion total parameters and 3 billion active parameters per token. It uses a hybrid sparse mixture-of-experts architecture combining Gated...
HAPPY9.8
96 mentions
SAD5.2
22 mentions
VIBE TRAJECTORY
26 weeks · 789 voices
HONEYMOON FADING
Early praise is cooling off in recent weeks.
HOW TO USE IT
Open weights. You can use a hosted service, or download it and run it yourself, free.
RUN IT ON YOUR OWN MACHINE
Open weights: download Qwen3.6 35B A3B once and it's yours. No subscription, no rate limits, works offline.
35B parameters (3B active per token), a Q4 file is about 18 GB
- BARE MINIMUM
NVIDIA RTX 4060 Ti 16GB
16 GB VRAM
~$390 NEWIQ3_S or Q3_K_S quant, ~15-20 tok/s, 16k context (requires partial CPU offload)
- SWEET SPOT
NVIDIA RTX 3090 24GB
24 GB VRAM
~$800 USEDIQ4_XS or Q4_K_M quant, ~30-45 tok/s, 32k context (fully in VRAM)
- FULL POWER
Apple Mac Studio M2 Ultra (128GB)
128 GB UNIFIED MEMORY
~$2,500 USEDBF16 (unquantized) or Q8_0, ~20-30 tok/s, up to 262k context
LEARN HOW TO SET IT UP
WILL MY BOX RUN QWEN3.6 35B A3B?
The three cards above are researched picks. Search the machine you actually have — we only say yes if it should reply at a usable speed, not just load the weights and crawl.
HARDWARE PICKS & PRICES ARE RESEARCHED FROM THE LIVE WEB AND REFRESHED AUTOMATICALLY. TREAT THEM AS BALLPARK, NOT GOSPEL.
WHAT CAN YOU ACTUALLY DO WITH IT?
Everyday jobs on each of those rigs: how long Qwen3.6 35B A3B takes, and how good the crowd says it is at that kind of work.
| THE JOB | BARE MINIMUMNVIDIA RTX 4060 Ti 16GB | SWEET SPOTNVIDIA RTX 3090 24GB | FULL POWERApple Mac Studio M2 Ultra (128GB) |
|---|---|---|---|
| Write an emaila paragraph or two, drafted from a one-line brief | ~29 S★★★★★ | ~13 S★★★★★ | ~20 S★★★★★ |
| Summarize a documenta long report boiled down to the points that matter | ~86 S★★★★★ | ~40 S★★★★★ | ~60 S★★★★★ |
| Build a websitea small landing page, markup and styles together | ~7.6 MIN★★★★★ | ~3.6 MIN★★★★★ | ~5.3 MIN★★★★★ |
| Build a backendan API with routes, storage and tests | ~24 MIN★★★★★ | ~11 MIN★★★★★ | ~17 MIN★★★★★ |
| Build a gamea playable browser game in one file | ~14 MIN★★★★★ | ~6.7 MIN★★★★★ | ~10 MIN★★★★★ |
STARS ARE THE COMMUNITY SCORE FOR THAT KIND OF WORK, DOCKED FOR HOW SQUASHED THE WEIGHTS GET AT EACH TIER. TIMES ARE COMPUTED FROM RESEARCHED SPEEDS. BALLPARK, NOT GOSPEL.
WHAT THE INTERNET SAYS
The internet is split on Qwen3.6 35B A3B. Biggest praise: help with code. Biggest gripe: saying no too much.
424 POSTS · 126 COMMENTS · REDDIT · HACKER NEWS · GITHUB · X · LEMMY · DEV FORUMS · BLOG
PEOPLE LOVE
- Popular for coding help (8 mentions)
- Answers come back fast (5 mentions)
- People feel it is worth the money (2 mentions)
PEOPLE COMPLAIN ABOUT
- Refuses or lectures more than people like (10 mentions)
- Outages, errors, or limits get in the way (6 mentions)
- Sometimes makes things up (3 mentions)
- Loses the thread in long chats (3 mentions)
TOPIC BY TOPIC
- Saying no too muchA SORE SPOT
0 thumbs up · 10 thumbs down
- Help with codePEOPLE LOVE THIS
8 thumbs up · 0 thumbs down
- Working when you need itA SORE SPOT
1 thumbs up · 6 thumbs down
- SpeedPEOPLE LOVE THIS
5 thumbs up · 1 thumbs down
- Getting facts rightMIXED FEELINGS
2 thumbs up · 3 thumbs down
- PriceMOSTLY POSITIVE
2 thumbs up · 1 thumbs down
- Long conversationsA SORE SPOT
0 thumbs up · 3 thumbs down
LOOKING FOR AN ALTERNATIVE?
Checked picks first: tags say why you would switch, stars say how fully each one stands in for Qwen3.6 35B A3B.
Muse Glimmer 30B
Grok 4.6> SOURCES (12)
- REDDIT · 2026-07-11 · 1420 UPVOTESQwen3.6 35B-A3B (Q8_0, no KV quant) single prompt in opencode: "Create a beautiful, relaxing flight simulator in a single html file with…
Qwen3.6 35B-A3B (Q8_0, no KV quant) single prompt in opencode: "Create a beautiful, relaxing flight simulator in a single html file with mountains, clouds, and endless procedural terrain"
- HACKER NEWS · 2026-04-16 · 1274 UPVOTESQwen3.6-35B-A3B: Agentic coding power, now open to all
Qwen3.6-35B-A3B: Agentic coding power, now open to all
- REDDIT · 2026-04-16 · 869 UPVOTESQwen3.6 is out now!
Qwen3.6 is out now!
- REDDIT · 2026-07-10 · 827 UPVOTES2.5x faster Qwen3.6 NVFP4 Unsloth quants
2.5x faster Qwen3.6 NVFP4 Unsloth quants
- REDDIT · 2026-05-18 · 507 UPVOTESRun Qwen3.6 MTP GGUFs locally!
Run Qwen3.6 MTP GGUFs locally!
- HACKER NEWS · 2026-04-16 · 463 UPVOTESQwen3.6-35B-A3B on my laptop drew me a better pelican than Claude Opus 4.7
Qwen3.6-35B-A3B on my laptop drew me a better pelican than Claude Opus 4.7
- REDDIT · 2026-04-30 · 318 UPVOTESLocal AI News You Missed - April 2026
Local AI News You Missed - April 2026
- REDDIT · 2026-04-16 · 311 UPVOTES2-bit Qwen3.6-35B-A3B GGUF is amazing! Made 30+ successful tool calls
2-bit Qwen3.6-35B-A3B GGUF is amazing! Made 30+ successful tool calls
- REDDIT · 2026-05-13 · 290 UPVOTESQwen3.6 MTP Unsloth Experimental GGUFs
Qwen3.6 MTP Unsloth Experimental GGUFs
- REDDIT · 2026-04-23 · 268 UPVOTESLast week in Generative Image & Video
Last week in Generative Image & Video
- REDDIT · 2026-08-04 · 267 UPVOTESA llama.cpp PR caches “hot” MoE experts on the GPU — 33 → 56 tok/s reported with 8GB VRAM
A llama.cpp PR caches “hot” MoE experts on the GPU — 33 → 56 tok/s reported with 8GB VRAM
- REDDIT · 2026-07-20 · 265 UPVOTESI ran Ternary-Bonsai-27B (2-bit) and Bonsai-27B (1-bit) on Terminal-Bench 2.0, in 8GB VRAM
I ran Ternary-Bonsai-27B (2-bit) and Bonsai-27B (1-bit) on Terminal-Bench 2.0, in 8GB VRAM