Did a bit of testing and thought I'd post the result here.

I tried this with Qwen using vLLM on 7900 XTX. The headline result was better than I expected, but NVFP4 still ended up being the limiting factor.

Happy to post more details if anyone wants them.

This is mostly for local work rather than public serving.

This keeps coming up in conversation so I figured I'd start a thread.

It feels like the conversation has shifted hard toward inference and serving details lately, and honestly that makes sense once you start dealing with tool calling.

Wondering if this is just where the whole space is heading now.

Thought I'd post a small recipe because a lot of threads ask similar things.

Maybe useful as a starting point: TensorRT-LLM on mini PC, keeping long context in mind from the start.

Maybe useful as a starting point for newer folks.

Posting this in case it saves someone else a little time.

I'm not trying to build a perfect architecture diagram. I just want a sane starting point around MLX and mini PC.

I'm especially interested in mistakes worth avoiding.

I'm trying to sanity-check my setup.

I'm trying to decide whether to keep the current stack or change course. Main question is how people here are handling small local stacks.

Not looking for the perfect answer, mostly what worked in real use.

I care more about stability than squeezing the last possible tokens/sec out of it.

Posting this in case it saves someone else a little time.

I'm helping a small team set up a local workflow. We don't need a giant stack, but we do care about VRAM headroom.

Would rather learn from your scars than collect my own all at once.

Posting this in case it saves someone else a little time.

If you were starting from scratch today for this use case, what would you choose first and what would you ignore?

I'm especially interested in mistakes worth avoiding.

Spent some time on this today and wanted to compare notes.

I'm comparing two approaches here: keep it simple on 5090, or squeeze more out of the stack and deal with the complexity.

Trying to avoid rebuilding the whole thing twice.

Thought I'd post a small recipe because a lot of threads ask similar things.

This isn't revolutionary, but the combination of PyTorch and M3 Max felt surprisingly usable once I tuned around prefill/decode split.

If someone wants the exact settings I can clean them up and post them.

For context, I'm not chasing leaderboard numbers. I mostly want the system to feel consistent day to day.

This keeps coming up in conversation so I figured I'd start a thread.

It feels like the conversation has shifted hard toward inference and serving details lately, and honestly that makes sense once you start dealing with tokens/sec.

Maybe obvious to some of you, but it clicked for me this week.