LOADING DATASHEET
GROUNDED IN REAL HUMAN CONVERSATIONSYOUR BENCH
Qwen2.5 72B Instruct
◆ OPEN
by Alibaba
OPEN WEIGHTSLOCALAPIPAIDSMALLCHINA-MADE
RANKED #292 OF 316 ASSISTANTS · OVERALL #451 OF 493 RANKED · VIBE SCORE 5.1 · 60 VOICES ·
COPY BADGE pastes a README snippet with the rank SVG.
Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2: - Significantly more knowledge and has greatly improved capabilities in coding and...
HAPPY7.3
7 mentions
SAD7.7
1 mentions
VIBE TRAJECTORY
26 weeks · 36 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 Qwen2.5 72B Instruct once and it's yours. No subscription, no rate limits, works offline.
72.7B parameters, a Q4_K_M GGUF is about 44 GB and FP16 weights require about 145 GB
- BARE MINIMUM
Mac mini M4 Pro (64 GB unified memory)
64 GB UNIFIED MEMORY
~$1,399 NEWIQ3_M or Q3_K_M quant, ~10 tok/s, 8k context
- SWEET SPOT
Dual NVIDIA RTX 3090 24GB rig
48 GB VRAM
~$1,500 USEDQ4_K_M or INT4 AWQ, ~18 tok/s, 16k context
- FULL POWER
Apple Mac Studio M2 Ultra (192 GB unified memory)
192 GB UNIFIED MEMORY
~$4,200 REFURBISHEDQ8_0 or FP16 unquantized, ~14 tok/s, 64k+ context
No live install guides found yet. The hardware picks above still stand.
WILL MY BOX RUN QWEN2.5 72B INSTRUCT?
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 Qwen2.5 72B Instruct takes, and how good the crowd says it is at that kind of work.
| THE JOB | BARE MINIMUMMac mini M4 Pro (64 GB unified memory) | SWEET SPOTDual NVIDIA 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 | ~50 S | ~28 S | ~36 S |
| Summarize a documenta long report boiled down to the points that matter | ~2.5 MIN★★★★★ | ~83 S★★★★★ | ~1.8 MIN★★★★★ |
| Build a websitea small landing page, markup and styles together | ~13 MIN★★★★★ | ~7.4 MIN★★★★★ | ~9.5 MIN★★★★★ |
| Build a backendan API with routes, storage and tests | ~42 MIN★★★★★ | ~23 MIN★★★★★ | ~30 MIN★★★★★ |
| Build a gamea playable browser game in one file | ~25 MIN★★★★★ | ~14 MIN★★★★★ | ~18 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 Qwen2.5 72B Instruct yet to call it. We found 61 posts but people did not say much either way.
41 POSTS · 20 COMMENTS · REDDIT · GITHUB · DEV FORUMS · LEMMY · OTHER FORUMS · HACKER NEWS · BLOG
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)
- HACKER NEWS · 2025-01-21 · 85 UPVOTESAsk HN: Can we just admit we want to replace jobs with AI?
Ask HN: Can we just admit we want to replace jobs with AI?
- LEMMY · 2025-01-30 · 76 UPVOTESDeepseek Tianenmen square controversy gets weirder
Deepseek Tianenmen square controversy gets weirder
- LEMMY · 2025-01-29 · 74 UPVOTESChinese firms ‘distilling’ US AI models to create rival products, warns OpenAI
Chinese firms ‘distilling’ US AI models to create rival products, warns OpenAI
- LEMMY · 2024-12-27 · 66 UPVOTESOpenAI announces plan to transform into a for-profit company
OpenAI announces plan to transform into a for-profit company
- REDDIT · 2024-09-17 · 44 UPVOTESQwen2.5-72B-Instruct is now on LMSys Chatbot Arena
Qwen2.5-72B-Instruct is now on LMSys Chatbot Arena
- BLOG · 2026-09-08 · 32 UPVOTESLatest open artifacts (#24): Motif-3, GLM-5.3, Hy4-preview and open model licenses
Avid Artifacts readers know that we have been covering not only models but also their licenses for quite some time. There was a period when custom licenses were all the rage, for example the custom Qwen2.5 72B-Instruct license or the Llama licenses . DeepSeek…
- REDDIT · 2026-04-29 · 13 UPVOTESTrying to build a local Claude Sonnet-like CLI coding setup on dual RTX 3090 — looking for model/backend/workflow advice
Trying to build a local Claude Sonnet-like CLI coding setup on dual RTX 3090 — looking for model/backend/workflow advice
- LEMMY · 2024-11-05 · 8 UPVOTESPerplexity CEO offers AI company's services to replace striking NYT staff TechCrunch
Perplexity CEO offers AI company's services to replace striking NYT staff TechCrunch
- LEMMY · 2026-09-01 · 7 UPVOTESA Strategic Coordination Framework of Small LMs Matches Large LMs in Data Synthesis
A Strategic Coordination Framework of Small LMs Matches Large LMs in Data Synthesis
- REDDIT · 2024-10-08 · 6 UPVOTESBest open-sourced coding LLM : Qwen2.5
Best open-sourced coding LLM : Qwen2.5
- REDDIT · 2025-07-26 · 6 UPVOTESRunning bnb-4bit on vLLM
Running bnb-4bit on vLLM
- REDDIT · 2025-12-18 · 0 UPVOTESToday I lost all my faith in China's capabilities in AI.
Today I lost all my faith in China's capabilities in AI.