LOADING DATASHEET
GROUNDED IN REAL HUMAN CONVERSATIONSYOUR BENCH
Qwen3 Next 80B A3B
◆ OPEN
by Alibaba
OPEN WEIGHTSLOCALAPILONG CONTEXTPAIDREASONING
RANKED #195 OF 316 ASSISTANTS · OVERALL #292 OF 493 RANKED · VIBE SCORE 5.9 · 101 VOICES ·
COPY BADGE pastes a README snippet with the rank SVG.
Qwen3-Next-80B-A3B-Thinking is a reasoning-first chat model in the Qwen3-Next line that outputs structured “thinking” traces by default. It’s designed for hard multi-step problems; math proofs, code synthesis/debugging, logic, and agentic..
HAPPY9.9
16 mentions
SAD5.1
2 mentions
VIBE TRAJECTORY
26 weeks · 75 voices in this window
HOLDING UP
Recent vibe is steady. No clear fade in the crowd.
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 Next 80B A3B once and it's yours. No subscription, no rate limits, works offline.
80B total parameters (sparse MoE activating 3B parameters), a Q4 GGUF is about 45 GB to 48 GB
- BARE MINIMUM
Desktop PC with AMD Ryzen 7 7700X, 64 GB DDR5 RAM, and RTX 3
12 GB VRAM + 64 GB RAM
~$900 USEDQ3_K_M or Q4_K_M with partial GPU offload and CPU/system RAM offloading, ~6 tok/s, 8k context
- SWEET SPOT
Dual NVIDIA GeForce RTX 3090 24GB Rig
48 GB VRAM
~$1,600 USEDQ4_K_M fully loaded in VRAM, ~35 tok/s, 32k context
- FULL POWER
Apple Mac Studio (M2 Ultra, 192 GB Unified Memory)
192 GB UNIFIED RAM
~$5,500 NEWQ8_0 or FP16 / BF16 unquantized, ~25 tok/s, 128k+ context
LEARN HOW TO SET IT UP
WILL MY BOX RUN QWEN3 NEXT 80B 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 Next 80B A3B takes, and how good the crowd says it is at that kind of work.
| THE JOB | BARE MINIMUMDesktop PC with AMD Ryzen 7 7700X, 64 GB DDR5 RAM, and RTX 3 | SWEET SPOTDual NVIDIA GeForce RTX 3090 24GB Rig | FULL POWERApple Mac Studio (M2 Ultra, 192 GB Unified Memory) |
|---|---|---|---|
| Write an emaila paragraph or two, drafted from a one-line brief | ~83 S★★★★★ | ~14 S★★★★★ | ~20 S★★★★★ |
| Summarize a documenta long report boiled down to the points that matter | ~4.2 MIN★★★★★ | ~43 S★★★★★ | ~60 S★★★★★ |
| Build a websitea small landing page, markup and styles together | ~22 MIN★★★★★ | ~3.8 MIN★★★★★ | ~5.3 MIN★★★★★ |
| Build a backendan API with routes, storage and tests | ~69 MIN★★★★★ | ~12 MIN★★★★★ | ~17 MIN★★★★★ |
| Build a gamea playable browser game in one file | ~42 MIN★★★★★ | ~7.1 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
Not enough discussion about Qwen3 Next 80B A3B yet to call it. We found 101 posts but people did not say much either way.
88 POSTS · 13 COMMENTS · LEMMY · REDDIT · GITHUB · DEV FORUMS · X · HACKER NEWS
PEOPLE LOVE
- No fan mail yet. The crowd is still warming up.
PEOPLE COMPLAIN ABOUT
- No complaints yet. The internet is still looking for one. Suspicious.
LOOKING FOR AN ALTERNATIVE?
Other models the crowd has fully reviewed, starting with text models like this one.
Muse Spark 1.3
DeepSeek V4 Flash> SOURCES (12)
- REDDIT · 2025-11-28 · 119 UPVOTESQwen3-Next Dynamic GGUFs out now!
Qwen3-Next Dynamic GGUFs out now!
- REDDIT · 2026-01-12 · 101 UPVOTESQwen3-Next-80B Instruct, Thinking Updated - 20% faster
Qwen3-Next-80B Instruct, Thinking Updated - 20% faster
- REDDIT · 2025-12-09 · 46 UPVOTESQwen3-Next here!
Qwen3-Next here!
- REDDIT · 2026-02-04 · 20 UPVOTESAlibaba releases Qwen3-Coder-Next: SWE-Bench 70.6, slightly above DeepSeek V3.2
Alibaba releases Qwen3-Coder-Next: SWE-Bench 70.6, slightly above DeepSeek V3.2
- LEMMY · 2025-11-29 · 18 UPVOTESQwen3-Next-80B-A3B Thinking and Instruct land in llama.cpp
Qwen3-Next-80B-A3B Thinking and Instruct land in llama.cpp
- REDDIT · 2025-09-25 · 11 UPVOTESTested Qwen3 Next on String Processing, Logical Reasoning & Code Generation. It’s Impressive!
Tested Qwen3 Next on String Processing, Logical Reasoning & Code Generation. It’s Impressive!
- REDDIT · 2025-09-25 · 11 UPVOTESTested Qwen3 Next on String Processing, Logical Reasoning & Code Generation. It’s Impressive!
Tested Qwen3 Next on String Processing, Logical Reasoning & Code Generation. It’s Impressive!
- LEMMY · 2025-11-29 · 8 UPVOTESQwen3-Next-80B-A3B Thinking and Instruct land in llama.cpp
Qwen3-Next-80B-A3B Thinking and Instruct land in llama.cpp
- REDDIT · 2025-12-04 · 7 UPVOTESGPT-OSS 120B vs QWEN3-NEXT 80B
GPT-OSS 120B vs QWEN3-NEXT 80B
- DEV FORUMS · 2025-11-19 · 4 UPVOTESSupport for RTX 6000 Blackwell 96GB card
Support for RTX 6000 Blackwell 96GB card
- REDDIT · 2026-04-11 · 0 UPVOTESWhy no talk about Medium (size) Language Models? 70-200B
Why no talk about Medium (size) Language Models? 70-200B
- GITHUB · 2026-09-05 · 0 UPVOTESCampaign: Atlas vs vLLM on NVIDIA datacenter GPUs (H100 / H200 / B200) — PRD, recipes, receipts
Campaign: Atlas vs vLLM on NVIDIA datacenter GPUs (H100 / H200 / B200) — PRD, recipes, receipts