SpanveroCheck my hardwareCompare GPUsModels by VRAMCompare costsHow it works

The best open LLMs you can run on 12 GB of VRAM

Open LLMs calculated to need no more than 12 GB of VRAM at the stated default quant and context — the RTX 3060 12 GB, 4070, or 6700 XT tier. Ranked by popularity within that calculated fit set, with local, rented-GPU, and API cost estimates. Verify the selected build and context on your exact hardware.

How this is ranked: Transparent calculated fit filter. 'Best' means 'estimated to run within a 12 GB VRAM budget at the stated quant and context.' Ordering is by popularity/recognition, not a quality benchmark; exact usage varies by runner and settings.

Showing the top 40 of 331. See all →

More: all "best" lists · Outcome Lab · all models

Open the free LLM cost calculator → · $0 markup. © 2026 Cynosure LLC.

The weekly price index

A short email of real AI price moves, straight from the daily log — no hype. We're collecting the list now; the first issue goes out when it opens. Unsubscribe with one click.