Skip to content
m.recipes
Find a configuration
Sign in

Run an AI model on your computer

Get a startup command, then explore measured performance for your task and GPU.

Get your startup command

Choose your operating system and enter a model repository. Launch support is checked separately from performance results.

All tasks · RTX 5090 · 32 GB

1 model with qualified results. Speeds compare only within one exact test; use “Same test on other GPUs” to compare hardware.

  1. Qwen3.8-27B-NVFP4

    RadixArk · 2 tested configurations

    • NVFP4 · 8K → 2K tokens
      Output throughput · includes prefill · 1 request
      70.8tok/s
      Observed GPU occupancy
      27.3GiB
      First token
      633 ms
      Use this recipe · 1 request ↗Evidence report and run availability ↗Same test on other GPUs
      Test details

      Total output tokens divided by the full request window, including prefill and queueing.

      Occupancy includes engine buffers and reserved KV cache. It is not the minimum VRAM required by the model.

      Reports combined counts reports, not inference requests. Request successes and failures are listed in each report.

      Engine
      vllm
      Protocol
      wbench-fixed-http-v1
      Dataset
      synthetic-tokenizer-v1:repeats=2
      Workload
      8192in-2048out
      ConcurrencyMean throughput (tok/s)P95TTFTObserved occupancy (GiB)GPU power (W)Reports combinedView report
      C170.8229.01 s633 ms27.29405.51bad294c0 ↗
    • 1 other test configurations on this GPU
      • NVFP4 · 32K → 2K tokens
        Output throughput · includes prefill · 1 request
        62.5tok/s
        Observed GPU occupancy
        27.3GiB
        First token
        3.6 s
        Use this recipe · 1 request ↗Evidence report and run availability ↗Same test on other GPUs
        Test details

        Total output tokens divided by the full request window, including prefill and queueing.

        Occupancy includes engine buffers and reserved KV cache. It is not the minimum VRAM required by the model.

        Reports combined counts reports, not inference requests. Request successes and failures are listed in each report.

        Engine
        vllm
        Protocol
        wbench-fixed-http-v1
        Dataset
        synthetic-tokenizer-v1:repeats=2
        Workload
        32768in-2048out
        ConcurrencyMean throughput (tok/s)P95TTFTObserved occupancy (GiB)GPU power (W)Reports combinedView report
        C162.4832.79 s3.6 s27.29433.31746bbc39 ↗

Models shown: 1 / 1

All results for this choice (1) Everything measured on this GPU

Recipe availability is checked on each report against the published recipe executor.

Occupancy includes engine buffers and reserved KV cache. It is not the minimum VRAM required by the model.

Measured hardware

13 more GPUs without qualified results for this task

GPU not listed? Test your own GPU

Running benchmarks

0Updates automatically

No benchmarks are currently reported as running.

RTX PRO 6000 Blackwell Server Edition · 96 GB · 403 queued · 0 running · 0 awaiting reassignment · 6 succeeded · 3 failed or exhausted

L4 · 24 GB · 91 queued · 0 running · 0 awaiting reassignment · 22 succeeded · 7 failed or exhausted

T4 · 16 GB · 22 queued · 0 running · 0 awaiting reassignment · 3 succeeded · 8 failed or exhausted