selfhost.directory

The index of self-hostable projects — ranked, categorized and served with live updates.
Own your data and stop giving it to big tech.

FINAL-Bench logo

Darwin-180B-RSI-R3

FINAL-Bench·Instruct·released 1 Oct 2026

qwen-community-1.0180BA61.6BMoE256K ctx ≈108 GB
Your GPU
ollama run hf.co/FINAL-Bench/POCKET-Darwin-180B-GGUF:UD-Q4_K_XL
All 6 runners ↓

💻 Run it on your own machine: POCKET-Darwin-180B-GGUF, the 4-bit GGUF of R3 (111 GB), runs on a laptop with an 8 GB GPU and 32 GB RAM, CPU only at 18–21 tok/s, a 128 GB mini PC or one DGX Spark. MMLU-Pro is identical to BF16 (87.65%).

Notify me when

How much VRAM does Darwin-180B-RSI-R3 need?

At Q4_K_M Darwin-180B-RSI-R3 needs about 108 GB with an 8K context (108 GB at Q4, 353 GB at full precision). Pick your GPU to see what fits and how fast it runs.

--vram
any
--gpus
--quant

Needs: ≈108 GB · Q4_K_M · 8K context — Pick a GPU to check the fit.

Change quant, context or KV cache to compare
Advanced: context 8K · KV cache 16-bit
Context
KV cache

Smallest GPU for each quant and context (KV cache 16-bit)

Quantisation8K16K32K64K128K256K
Q3_K_M
Q4_K_M
Q5_K_M
Q6_K
Q8_0
FP16

≤16 GB card24–32 GB48–96 GBmore

QuantisationQualityWeightsKV cacheTotal
Q2_KHeavy lossSmallest, noticeably worse qualityHeavy loss76 GB file0.8 GB≈81 GB
Q3_K_MNoticeable lossSmall, some quality lossNoticeable loss89 GB file0.8 GB≈95 GB
Q4_K_MrecommendedSmall lossRecommended default for local useSmall loss102 GB est.0.8 GB≈108 GB
Q5_K_MNear-losslessNear-lossless for most tasksNear-lossless119 GB est.0.8 GB≈126 GB
Q6_KNear-losslessPractically losslessNear-lossless138 GB est.0.8 GB≈146 GB
Q8_0LosslessLossless in practice, twice the size of Q4Lossless178 GB est.0.8 GB≈188 GB
FP16 / BF16LosslessFull precision, as releasedLossless335 GB est.0.8 GB≈353 GB

Estimates for 8K tokens of context: weights + KV cache + runtime overhead. All experts stay in memory, but only the active ones are read per token. Speed assumes memory-bound decoding at your GPU’s bandwidth. Pick a GPU above and the fit column checks every row against it. Weights marked “file” use the real GGUF download size; “est.” rows are computed from the parameter count.

How to pick a quant

Take the biggest quant that fits with room left for context. Q4_K_M is the usual sweet spot. Below Q4 quality drops fast, above Q6 you mostly pay memory for nothing. Big models (70B+, large MoE) tolerate low quants better than small ones.

  • LosslessF16, Q8_0
  • Near-losslessQ6_K, Q5_K_M
  • Small lossQ5_K_S, Q4_K_M, IQ4_NL, IQ4_XS, Q4_K_S
  • Noticeable lossQ4_0, Q3_K_L, Q3_K_M, IQ3_M
  • Heavy lossIQ3_XXS, Q3_K_S, IQ2_M, Q2_K

Typical quality loss, not measured per model. Tiers come from KL divergence against FP16 on Llama 3 8B: llama.cpp perplexity README: LLaMA 3 8B scoreboard (KLD and ΔPPL vs FP16, Wikitext-2); llama.cpp quantize tool: ΔPPL per quant type @ Llama-3-8B

How much context fits?

Model quant
KV cache

At 8K context, Q4_K_M needs ≈108 GB(102 GB weights + 0.8 GB KV cache + 5.6 GB overhead). Pick a GPU above to see how much context fits.

0501001501K2K4K8K16K32K64K128K256KGB

Weights + overheadKV cache

Can the Apple M2 Ultra (192 GB) run Darwin-180B-RSI-R3? →

Speed vs quality per quant

Smaller quants run faster and fit in less memory, but lose quality.

Q2_KQ3_K_MQ4_K_MQ5_K_MQ6_KQ8_0F16

Fits in VRAMTight fitCPU offload✕ won't runQuality is typical KL divergence vs FP16, not measured for this model.

Real-world speed of Darwin-180B-RSI-R3

Tokens per second measured by people running it at home. A GPU shows a median once 3 reports agree on the same setup. Reports far from our estimate wait for a review. Measure yours with llama-bench →

No medians yet. Be one of the first 3 to report a setup.

All reports

Run Darwin-180B-RSI-R3 locally

--quantQ4_K_Mdefault; pick your GPU to tailor itFINAL-Bench/POCKET-Darwin-180B-GGUF

UD-Q4_K_XL straight from Hugging Facedocs
ollama run hf.co/FINAL-Bench/POCKET-Darwin-180B-GGUF:UD-Q4_K_XL

Install Ollama

OpenAI-compatible server with a web UIdocs
llama-server -hf FINAL-Bench/POCKET-Darwin-180B-GGUF:UD-Q4_K_XL

Downloads the GGUF on first run. Installed through llama.app? The same flags work with `llama serve`.

Install llama.cpp

Download with the lms CLIdocs
lms get FINAL-Bench/POCKET-Darwin-180B-GGUF@UD-Q4_K_XL

Then load it from the LM Studio app or with `lms load`.

Install LM Studio

Serve an OpenAI-compatible APIdocs
vllm serve FINAL-Bench/Darwin-180B-RSI-R3
Docker Compose (NVIDIA)docs
yaml
services:
  vllm:
    image: vllm/vllm-openai:latest
    command: FINAL-Bench/Darwin-180B-RSI-R3
    ports:
      - "8000:8000"
    ipc: host
    volumes:
      - ~/.cache/huggingface:/root/.cache/huggingface
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]

API on port 8000.

Install vLLM

Serve an OpenAI-compatible APIdocs
sglang serve FINAL-Bench/Darwin-180B-RSI-R3 --host 0.0.0.0 --port 30000
Docker Compose (NVIDIA)docs
yaml
services:
  sglang:
    image: lmsysorg/sglang:latest
    command: sglang serve FINAL-Bench/Darwin-180B-RSI-R3 --host 0.0.0.0 --port 30000
    ports:
      - "30000:30000"
    ipc: host
    volumes:
      - ~/.cache/huggingface:/root/.cache/huggingface
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]

API on port 30000.

Install SGLang

Ask about an imagedocs
python
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="FINAL-Bench/Darwin-180B-RSI-R3", dtype="auto", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/bee.jpg"},
            {"type": "text", "text": "Describe this image."},
        ],
    }
]
out = pipe(text=messages, max_new_tokens=256)
print(out[0]["generated_text"][-1]["content"])

Install Transformers

Quantised versions and downloads

About Darwin-180B-RSI-R3

💻 Run it on your own machine: POCKET-Darwin-180B-GGUF, the 4-bit GGUF of R3 (111 GB), runs on a laptop with an 8 GB GPU and 32 GB RAM, CPU only at 18–21 tok/s, a 128 GB mini PC or one DGX Spark. MMLU-Pro is identical to BF16 (87.65%). reasoning · MoE 512 experts · 262K long context · image + text · Korean + English · self-improvement · structured output · ZTC The second round of model-level self-improvement on top of Darwin-180B-RSI. R3 continues training from R1 with the same recipe: the model solves verifiable problems, keeps only its own solutions that check out as correct, and trains on them. No human-written solutions or reasoning traces are used. R3 is released so anyone can download it, run it, and check the numbers below. --- Hugging Face official benchmark leaderboards. Each score is listed under the model that produced it. R3's own leaderboard entries: Full settings are recorded in .eval_results/. Scores as listed on the Hugging Face official benchmark leaderboards (self-reported by each model's publisher). "·" = not reported. Open-weight models only; closed API models are not included. Settings (samples, voting, thinking budget) differ across models. R1 settings are in the evaluation protocol below. ---

Full model card on Hugging Face →

Alternatives to Darwin-180B-RSI-R3

All alternatives →
Qwen logo

Qwen3.8-Flash-Next

Qwen·Vision

≈108 GB

As the frontier of foundation models pushes toward ever-larger parameter counts and ever-longer context windows, the question is no longer just how much we can scale, but how efficiently we can do so.

180BA61.6B256K ctx qwen-community-1.0
2.7MAug 2026ollama
MiniMaxAI logo

MiniMax-M3.1-Flash

MiniMaxAI·Vision

≈132 GB

MiniMax-M3.1-Flash is a native multimodal mixture-of-experts model for coding, agentic workflows, computer use, and cybersecurity. It supports text, image, and video inputs with a context window of up to 1M tokens.

220B1M ctx minimax-community
150KOct 2026ollama
zai-org logo

GLM-5.3-Flash

zai-org·Vision

≈213 GB

👋 Join our WeChat or Discord community. 📖 Check out the GLM-5.3-Flash blog and GLM-5 Technical report . 📍 Use GLM-5.3-Flash API services on Z.ai API Platform.

321BA18.3B1M ctx MIT
6.4MAug 2026ollama
thinkingmachines logo

Inkling-Small

thinkingmachines·Vision

≈159 GB

Inkling-Small is a general-purpose multimodal model that accepts text, image and audio inputs and generates text outputs. It is intended for use in English and other languages, and across multiple coding languages.

266BA1.7B1M ctx Apache 2.0
916.1KJul 2026gguf
Qwen logo

Qwen3.5-122B-A10B

Qwen·Vision

≈75 GB

Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.

125BA10B256K ctx Apache 2.0
706.9KFeb 2026ollama

NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16 is the full-precision release of NVIDIA Nemotron 3.5 Super VL, a vision-language model for reasoning, coding, tool use, document understanding, and video understanding.

124BA12B1M ctx openmdw-1.1
82.2KOct 2026gguf

Darwin-180B-RSI-R3 FAQ

How much VRAM does Darwin-180B-RSI-R3 need?
About 108 GB at Q4_K_M with an 8K context. Q8_0 needs about 188 GB and full precision about 353 GB. Longer contexts need more memory for the KV cache.
Can Darwin-180B-RSI-R3 run on the Apple M2 Ultra (192 GB)?
Yes. It fits at Q5_K_M (about 126 GB), at an estimated 10 tokens/s. It fits comfortably on the Apple M3 Ultra (512 GB).
Can I use Darwin-180B-RSI-R3 commercially?
Yes, with conditions. It is released under qwen-community-1.0. Custom licence: read the model card before using it commercially.
What is the context length of Darwin-180B-RSI-R3?
262,144 tokens (256K). The full window needs far more memory than an 8K context.
How do I run Darwin-180B-RSI-R3 with Ollama or llama.cpp?
With Ollama: ollama run hf.co/FINAL-Bench/POCKET-Darwin-180B-GGUF:UD-Q4_K_XL. With llama.cpp: llama-server -hf FINAL-Bench/POCKET-Darwin-180B-GGUF:UD-Q4_K_XL, which starts an OpenAI-compatible server.

Data from Hugging Face, the model card and public leaderboards. Memory and speed figures are estimates; real numbers depend on the runtime, context length and drivers. Always check the licence yourself before commercial use.