boltbeam.fyi

Make your inference engine faster, or you don't pay

I make inference engines faster on your hardware: open-source engines like llama.cpp, vLLM and TensorRT-LLM, or your own closed-source one. You get a fixed quote with a speed target written in. If the target is not met, you pay nothing.

  1. 1Your setup

    Tap what fits. Skip what you don't know.

    GPU
    How many GPUs
    System RAM
    CPU
    Inference engine
    Model size
    Workload
    Context
    What should get faster
    Anything else

    0 of 9 answered

  2. 2Pick a time for the free audit

    The calendar opens soon. Until then, email julianabeleda@arkey.ai with your setup and I'll send times.
  3. 3Save a card: $0 today

    Nothing is charged now. Your card is saved with Stripe, and charged only if two things happen: you accept a quote, and the speed test in it passes. Miss the target, and it is never charged.

    Payments by Stripe. boltbeam.fyi never sees your card number.

How it works

  1. You tell me the setup: engine, model, GPU, and the workload you care about (one user or many, short or long context).
  2. I measure it first, free. BoltBeam audits your setup against the roofline: the fastest your GPU can run that model. The gap is what can be won, and it tells both of us whether the job is worth doing.
  3. You get a fixed quote with one test in it. For example: decode at least 15% faster at batch 1 on your workload, measured by BoltBeam, before and after, paired runs on the same machine.
  4. I do the work. Engine patches, kernels, settings, or a lighter checkpoint, whichever closes the gap.
  5. The test decides. Same benchmark, same machine. If it passes, the quote is charged. If it does not, nothing is.

The guarantee

  • Card saved at $0 when you book. Nothing is charged for the audit.
  • Target met: the quoted amount is charged.
  • Target missed: nothing is charged, and you keep the audit.

Open or closed source

Open-source engines: the work can go upstream or stay as patches you own, your choice. Closed-source engines: fine, with your written permission to profile and modify the engine, under an NDA. The work runs on your hardware or a rented machine you approve, and the results stay private unless you want them published.

What I have shipped