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.

microsoft logo

Can the RTX 4070 run UserLM-2.0?

UserLM-2.0·8B·GeForce RTX 4070·12 GB VRAM

Saved in this browser, used on every model page.
Yes

Yes. The RTX 4070 runs UserLM-2.0 fully in VRAM at Q8_0 (about 9.8 GB of 12 GB), at roughly 32 tokens/s.

best quant
Q8_0
needs
9.8 GB
usable
12 GB
est. speed
~32 tok/s

Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the UserLM-2.0 page →

Every quantisation of UserLM-2.0 on the RTX 4070

QuantQualityMemoryFitSpeed
Q2_K Heavy loss4.1 GBfits~105 tok/s
Q3_K_M Noticeable loss5.3 GBfits~71 tok/s
Q4_K_M Small loss6.3 GBfits~57 tok/s
Q5_K_M Near-lossless7.1 GBfits~49 tok/s
Q6_K Near-lossless8.0 GBfits~42 tok/s
Q8_0 pickLossless9.8 GBfits~32 tok/s
FP16 / BF16 Lossless17 GBoffload~5.3 tok/s

Speed is a bandwidth-based estimate for generating text (504 GB/s); prompt processing and long contexts are slower.

Measured speed on the RTX 4070

No community reports for this pair yet.

Report your own speed →

Run UserLM-2.0 on the RTX 4070

Using Q8_0, the best fit for RTX 4070.

llama.cpp · OpenAI-compatible server with a web UIdocs
llama-server -hf liodon-ai/UserLM-2.0-imatrix-GGUF:Q8_0

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

vLLM · Serve an OpenAI-compatible APIdocs
vllm serve microsoft/UserLM-2.0
vLLM · Smaller FP8 builddocs
vllm serve liodon-ai/UserLM-2.0-FP8
vLLM · Docker Compose (NVIDIA)docs
services:
  vllm:
    image: vllm/vllm-openai:latest
    command: microsoft/UserLM-2.0
    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.

SGLang · Serve an OpenAI-compatible APIdocs
sglang serve microsoft/UserLM-2.0 --host 0.0.0.0 --port 30000
SGLang · Docker Compose (NVIDIA)docs
services:
  sglang:
    image: lmsysorg/sglang:latest
    command: sglang serve microsoft/UserLM-2.0 --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.

Transformers · Chat with the pipeline APIdocs
from transformers import pipeline

pipe = pipeline("text-generation", model="microsoft/UserLM-2.0", dtype="auto", device_map="auto")
messages = [{"role": "user", "content": "Explain self-hosting in one paragraph."}]
out = pipe(messages, max_new_tokens=512)
print(out[0]["generated_text"][-1]["content"])

Chat models that fit the RTX 4070 comfortably

All that fit →
ModelParamsBest quantMemorySpeedLicence
8BQ8_0≈9.8 GBFits in memory~32 tok/sConditions
8.2BQ8_0≈10 GBFits in memory~32 tok/sCommercial OK
8BQ8_0≈9.8 GBFits in memory~32 tok/sConditions
7.6BQ8_0≈8.9 GBFits in memory~34 tok/sCommercial OK
9.4BQ8_0≈11 GBFits in memory~28 tok/sCommercial OK
7.2BQ8_0≈9.0 GBFits in memory~36 tok/sCommercial OK

Other GPUs that run UserLM-2.0

Frequently asked questions

Can the RTX 4070 run UserLM-2.0?
Yes. The RTX 4070 runs UserLM-2.0 fully in VRAM at Q8_0 (about 9.8 GB of 12 GB), at roughly 32 tokens/s.
How fast does UserLM-2.0 run on the RTX 4070?
An estimate from memory bandwidth gives about 32 tokens/s at Q8_0.
Which quantisation of UserLM-2.0 should I use on the RTX 4070?
Q8_0: the largest one that fits. Expected quality: lossless compared with the full-precision weights.

Memory and speed are estimates from the model's size and the card's memory bandwidth; real numbers depend on the runtime, drivers and context length. Community figures are medians of user reports.