LOADING DATASHEET
GROUNDED IN REAL HUMAN CONVERSATIONSYOUR BENCH
R1
◆ OPEN
by DeepSeek
OPEN WEIGHTSLOCALAPIPAIDCLOSEDREASONING
RANKED #83 OF 228 ASSISTANTS · OVERALL #140 OF 6,618 · VIBE SCORE 6.1 · 103 VOICES
May 28th update to the [original DeepSeek R1](/deepseek/deepseek-r1) Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active...
HAPPY9.9
36 mentions
SAD5.1
11 mentions
VIBE TRAJECTORY
26 weeks · 30 voices
AGING WELL
The crowd is warmer now than it was at the start.
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 R1 once and it's yours. No subscription, no rate limits, works offline.
671B parameters MoE, a standard Q4 quant file is about 404 GB (distills exist from 1.5B to 70B)
- BARE MINIMUM
Mac Studio M2 Ultra (192GB Unified Memory)
192 GB UNIFIED MEMORY
~$5000 USEDExtreme low-bit quant (UD-IQ1_S or dynamic 1.58-bit) at ~130 GB footprint; the full 671B model is too large for standard
- SWEET SPOT
Dual Mac Studio M2 Ultra (192GB Unified Memory each) cluster
384 GB UNIFIED MEMORY
~$10000 USEDQ4_K_M quant split via llama.cpp or MLX distributed, comfortable reasoning inference at 16k context
- FULL POWER
Workstation with 8x NVIDIA RTX 3090 (24GB VRAM)
192 GB VRAM PLUS 512 GB SYSTEM RAM
~$12000 USEDK-Quant or FP8 hybrid offload across 8 GPUs with vLLM / SGLang, fast full-context inference
LEARN HOW TO SET IT UP
WILL MY BOX RUN R1?
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 R1 takes, and how good the crowd says it is at that kind of work.
| THE JOB | BARE MINIMUMMac Studio M2 Ultra (192GB Unified Memory) | SWEET SPOTDual Mac Studio M2 Ultra (192GB Unified Memory each) cluster | FULL POWERWorkstation with 8x NVIDIA RTX 3090 (24GB VRAM) |
|---|---|---|---|
| Write an emaila paragraph or two, drafted from a one-line brief | ~2.1 MIN★★★★★ | ~36 S★★★★★ | ~20 S★★★★★ |
| Summarize a documenta long report boiled down to the points that matter | ~6.3 MIN★★★★★ | ~1.8 MIN★★★★★ | ~60 S★★★★★ |
| Build a websitea small landing page, markup and styles together | ~33 MIN★★★★★ | ~9.5 MIN★★★★★ | ~5.3 MIN★★★★★ |
| Build a backendan API with routes, storage and tests | ~1.7 HR★★★★★ | ~30 MIN★★★★★ | ~17 MIN★★★★★ |
| Build a gamea playable browser game in one file | ~63 MIN★★★★★ | ~18 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
People are mostly happy with R1. Biggest praise: help with code.
78 POSTS · 25 COMMENTS · REDDIT · BLUESKY · DEV.TO · LEMMY · GITHUB · BLOG
PEOPLE LOVE
- People feel it is worth the money (4 mentions)
- Popular for coding help (5 mentions)
PEOPLE COMPLAIN ABOUT
- People find it expensive (1 mentions)
TOPIC BY TOPIC
- PricePEOPLE LOVE THIS
4 thumbs up · 1 thumbs down
- Help with codePEOPLE LOVE THIS
5 thumbs up · 0 thumbs down
LOOKING FOR AN ALTERNATIVE?
Other models the crowd has fully reviewed, starting with text models like this one.
GLM 5.3> SOURCES (12)
- REDDIT · 2025-07-23 · 529 UPVOTESChina is winning the AI race for coding while being open source
China is winning the AI race for coding while being open source
- REDDIT · 2025-05-28 · 395 UPVOTESDeepSeek-R1-0528
DeepSeek-R1-0528
- REDDIT · 2025-05-28 · 372 UPVOTESNEW DeepSeek-R1-0528 🔥 Let it burn
NEW DeepSeek-R1-0528 🔥 Let it burn
- REDDIT · 2025-05-30 · 355 UPVOTESYou can now run the full DeepSeek-R1-0528 model locally!
You can now run the full DeepSeek-R1-0528 model locally!
- REDDIT · 2025-05-30 · 343 UPVOTESYou can now run DeepSeek-R1-0528 on your local device! (20GB RAM min.)
You can now run DeepSeek-R1-0528 on your local device! (20GB RAM min.)
- REDDIT · 2025-05-29 · 197 UPVOTESDeepSeek R1 0528 just dropped today and the benchmarks are looking seriously impressive
DeepSeek R1 0528 just dropped today and the benchmarks are looking seriously impressive
- REDDIT · 2025-05-28 · 176 UPVOTESNew DeepSeek model released on Hugging Face DeepSeek-R1-0528
New DeepSeek model released on Hugging Face DeepSeek-R1-0528
- REDDIT · 2025-05-30 · 111 UPVOTESYou can now run DeepSeek R1-0528 locally!
You can now run DeepSeek R1-0528 locally!
- REDDIT · 2025-08-04 · 83 UPVOTESPerplexity removes the reasoning model R1, claiming it is an outdated model!!
Perplexity removes the reasoning model R1, claiming it is an outdated model!!
- REDDIT · 2025-05-29 · 83 UPVOTESDeepSeek R1 0528 just dropped today and the benchmarks are looking seriously impressive
DeepSeek R1 0528 just dropped today and the benchmarks are looking seriously impressive
- REDDIT · 2025-06-06 · 76 UPVOTESWhich APIs do you use for FREE - Best free options for CODING
Which APIs do you use for FREE - Best free options for CODING
- REDDIT · 2025-12-26 · 76 UPVOTESLiquid AI released an experimental checkpoint of LFM2-2.6B using pure RL, making it the strongest 3B on the market
Liquid AI released an experimental checkpoint of LFM2-2.6B using pure RL, making it the strongest 3B on the market