Home / Models / UI-Mate-27B

How to Run UI-Mate-27B Locally — Setup Guide (September 2026)

UI-Mate is Tencent's computer-use model family — vision models trained to look at a screenshot and decide where to click. That is a different job from general chat, and it is one of the few tasks where running locally is not just cheaper but structurally better: the screenshots never leave your machine. The 27B fits a 16 GB card; the 9B fits 8 GB.

Numbers you can trust

We don't print guessed speeds. Every measured number on this page came from TurboLLM's own auto-tuner running UI-Mate-27B on real hardware — labeled as such. When you run it yourself, TurboLLM benchmarks it on your exact GPU and shows a VRAM-fit verdict before you load, and the real measured tokens/sec once it's running.

Does it fit your GPU?

27.4B dense parameters, Apache-2.0, on the same hybrid-attention stack as current Qwen models — 16 of 64 layers keep a growing KV cache — with a 262K native context. Screen understanding needs the mmproj file alongside the weights.

UI-Mate-27B
Tencent · GGUF
Dense · hybrid linear attention · vision
visiontools
VRAMFit
16 GBQ3_K_M (14.4 GB) or IQ4_XS (15.3 GB) fully in VRAM alongside the mmproj · 262K context
24 GBQ4_K_M (17.5 GB) or Q5_K_M (20.5 GB) fully in VRAM · 262K context
32 GB+Q6_K (23.2 GB) or Q8_0 (28.7 GB) fully in VRAM · 262K context
bartowski/tencent_UI-Mate-27B-GGUF
View on Hugging Face →
UI-Mate-9B
Tencent · GGUF
Dense · vision
visiontools
VRAMFit
8 GBQ4_K_M (5.9 GB) fully in VRAM with the mmproj alongside it · 262K context
12 GB+Q6_K (7.7 GB) or Q8_0 (9.6 GB) fully in VRAM · 262K context
bartowski/tencent_UI-Mate-9B-GGUF
View on Hugging Face →

How to run it

  1. Install TurboLLM

    npx turbollm — no install step, works on Windows, macOS, and Linux. It detects your GPU and auto-provisions a matching llama-server build (no CUDA toolkit, no Python, no compiler).

  2. Get the model

    Paste bartowski/tencent_UI-Mate-27B-GGUF into TurboLLM's in-app Hugging Face search — it lists every quant with a VRAM-fit verdict against your real free VRAM, and downloads are resumable and SHA-256 verified.

  3. Let auto-fit pick the setup

    Flip the auto-fit toggle and TurboLLM chooses the GPU/CPU layer split (and, for a Mixture-of-Experts model, the expert-offload split) for your card — the same manual flag-hunting every other guide walks you through by hand.

  4. Load it

    TurboLLM benchmarks the model on your exact GPU as it loads and shows the real measured tokens/sec — not a number copied from someone else's card.

  5. Use it

    Chat in the built-in UI, or point any OpenAI- or Anthropic-compatible tool (including Claude Code) at TurboLLM's local API — see the API docs.

One honest caveat

A computer-use model is not a chat model. UI-Mate is trained to emit click coordinates and UI actions from a screenshot, so it will feel oddly terse if you talk to it like a general assistant — it wants a screen and a goal. Point an agent framework at TurboLLM's local API to use it properly.

FAQ

Will UI-Mate-27B run on my GPU?

Check the VRAM table above for the tier closest to your card. If you're not sure which tier you're in, or want picks tailored to your exact hardware, use What can I run? — enter your VRAM (or Apple unified memory) and it suggests the model, quant, and offload split that fits, the same logic TurboLLM's auto-fit runs when you load a model.

Which quant should I use for UI-Mate-27B?

Pick the highest quant that fits your VRAM with room left over for the KV cache and compute buffer — not just the model weights. As a starting point, Q4_K_M is usually the sweet spot (small quality loss for a large size drop); step up to Q5_K_M/Q6_K/Q8_0 if you have headroom, or down to Q3_K_M/an IQ quant if you don't. TurboLLM shows a VRAM-fit verdict for every quant against your card's real free VRAM before you download — see Quantization explained for what the letters and numbers actually mean, or llama.cpp flags that actually matter for the other settings worth knowing.

Run it now

$ npx turbollm

One command detects your GPU, provisions the right engine, and opens the UI. New here? Start with Install & first run and Quantization explained. On different hardware, see RTX 5070 Ti or RTX 3090. Also see how to run it or how to run it.