Can a Strix Halo Framework Desktop, 32GB run local LLMs?

Yes — 29 of the 48 current models on this site fit in its 24 GB of usable memory, and one more if you keep the window shorter than 32k. The strongest it holds, Qwen3.8 27B, runs at 10 tok/s with a 32k window, worked out from memory bandwidth. Whether that saves you money is a different question: at $1,269 and 500k tokens a day, it pays back in 91 years.

Run the numbers on this machine

What the Framework Desktop, 32GB runs

They do not all hold the same window. The weights are a fixed size, but the key-value cache grows with every token you keep, so what is left of 24 GB after the weights is how far the context goes. On 6 of the 12 below, this machine's memory is what runs out first: Gemma 4 31B it stops soonest, at 32k. The rest reach 256k, where the calculator's list ends, or the longest setting their own context limit allows.

Speed and memory are at 32k context, the setting the calculator starts on; the memory column is the weights plus the cache for that much of it. Each speed says how it was arrived at, and every one here is estimated from this machine's 256 GB/s of memory bandwidth rather than taken from a published benchmark. There is a page on how that sum works, and what each size needs, and one on what 32 GB runs, machine by machine. The longest context is the longest setting the calculator offers that this machine still holds the model at, cache included, and each one opens the calculator on that model at that length. A figure tagged memory is one this machine ran out of room for, and the rest are stopped by the model's own limit or by the end of the list.

For scale, the calculator starts from 80 tok/s for a hosted API and times a local machine against it. Nothing above reaches it, and the quickest is 65 tok/s.

17 more fit. Ranked below the twelve above: Ling 3.0 tiny, Granite 4.2 8B, Qwen3-Coder 30B-A3B, gpt-oss-20b, Gemma 4 E4B, LFM2.5 2.6B, Devstral Small 2 24B, Ministral 3 14B, Ministral 3 8B. The other 8 have no intelligence-index score, so the ranking cannot place them: Laguna XS 2.1, LFM3 1.7B, Ornith 1.5 9B, Ornith 1.5 35B-A3B, Spark-X2.5 4B, Spark-X2.6 4B, KAT-Coder V2.5 Dev 35B-A3B, LFM2.5 24B-A2B. Their pages show what each one needs and what runs it.

The 48 counted above are the current models. This site prices 16 superseded ones as well, because people still run them, and this machine holds eight of those at 32k, strongest first: Qwen3 30B-A3B Instruct 2507, Llama 3.1 8B Instruct at Q4_K_M, Llama 3.1 8B Instruct at Q8_0, Mistral Small 3.2 24B Instruct, Qwen3 14B, Qwen3 8B, Gemma 3 27B it, Gemma 3 12B it. Each of those pages asks whether the model is still worth running.

One more, at a shorter window

The table above counts what fits at 32k, the context the calculator starts on. One more model fits this machine at a shorter one. Same weights, smaller cache.

Each window is the longest setting the calculator offers that this machine still holds the model at, and it opens the calculator on that model at that length. The memory figures are the weights plus the key-value cache at that window, against the 24 GB this machine's GPU can use.

Other machines to weigh against the Framework Desktop, 32GB

Every row uses the same defaults as the figures above: 500k tokens a day at 15:1 input to output, 32k context, and each machine's strongest model that fits, counted against the same 48 current models. Graphics cards are priced as the card alone, so add the PC around one before comparing it with a complete computer. All seven cards here are ranked by what each one holds.

Head to head with another computer: Mac mini M6, 32GB. With the same box and the bigger chip: 64GB. There are 111 machine match-ups on the site.

The specifics

Chip

Framework Desktop — Ryzen AI Max 385 · Radeon 8050S, 32 CU

Memory bandwidth

256 GB/s — Measured GPU bandwidth is ~212-215 GB/s of the 256 GB/s theoretical, so estimates here run a little high.

Usable by the GPU

24 GB — Windows lets you dedicate up to 75% of RAM to the GPU (Variable Graphics Memory in AMD Adrenalin; Framework documents 96 GB on a 128 GB board). On Linux the amdgpu GTT pool can be raised to roughly 108-120 GB with ttm.pages_limit / amdgpu.gttsize boot parameters. The Windows figure is used here as the conservative one.

Power under load

133 W (stand-in) — No measurement for the Max 385 board; stand-in is the Max+ 395 board's measured 133 W, which likely overstates this smaller GPU.

Availability

System $1,269 (out of stock when checked); mainboard alone $969. Note: the 32 GB tier is the smaller Ryzen AI Max 385 with a 32-CU GPU.

Sources

Framework (configurator), Framework