# Choosing hardware for a local model · Good AI Labs

> reflex-1 runs on a laptop CPU as well as an H200. The request, latency target and runtime memory determine which machine fits the application.

URL: https://www.goodailabs.com/blog/sovereign-intelligence/
Date: 2026-10-02

reflex-1 took 345.2 ms median to select among 77 intents on one M4 CPU
thread. A separate H200 test returned recorded agent requests in 27.5 ms
median, including local HTTP. A developer can run the model on either;
the application's latency target determines which result is relevant.

The requests differ between these tests. To choose hardware for a service,
measure its actual input lengths and option counts at the expected
concurrency. The [reflex-1 report](/research/reflex-1/)
records the conditions for our measurements.

## Disk, memory and adaptation

The FP32 reflex-1 model occupies about 1.69 GB on disk and contains 421M
serving parameters. Its main adaptation run trained 9.04M parameters.
Storage, inference and adaptation therefore have different budgets.
Activations and framework allocations add to runtime memory, and the
number of simultaneous requests can change the peak.

Samsara illustrates the training cost. Its 217M-parameter policy was trained
on eight H200s. Near the end of the run, rank zero recorded 23.52 GiB peak
allocated memory. That training measurement includes costs such as gradients
and optimizer state; it does not specify inference requirements.
The [Samsara report](/research/samsara-vla/) gives the setup.

## The rest of the application

Retrieval, tool execution and logging also need hardware. Keeping inference
local controls where model inputs are processed, but each connected service
has its own data path and resource use.



Measure the full request from arrival to completed action. Model latency
is one part of that interval; queues and tool calls may account for more.

Contact: research@goodailabs.com