RX 9060 XT
An RX 9060 XT runs 12 of the 34 open models tracked here at Q4_K_M with 8k context, computed on 16 October 2026. The largest is gpt-oss 20B, needing 13.4 GB and generating at most 107 tokens per second.
- Memory
- 16 GB VRAM
- Bandwidth
- 320 GB/s
- Board power
- 160 W
- Released
- 2025
- Availability
- Sold new
Open the RX 9060 XT in the LLM checker
What an RX 9060 XT runs, model by model
RX 9060 XT 16 GB addresses 16 GB at 320 GB/s. Every model below is at Q4_K_M with 8k of context, the setting most people actually use. Once the 5% safety margin comes off, the card has 15.2 GB to spend: memory decides what loads, bandwidth how fast it answers. Both are estimates from stated formulas, and the VRAM calculator shows the working.
On this card a model fits at or under 15.2 GB, and above 13.6 GB it is tight: it loads, but there is little left for the conversation, and a desktop on the same 16 GB can push it over. Tokens per second are single-stream estimates from memory bandwidth; ≤ marks a ceiling real runtimes fall well below.
Computed 16 October 2026The model to actually run on it
Gemma 4 12B is the best use of this card: 8.2 GB of the 16 GB available, an estimated 32 tokens per second, comfortable for chat, and 131,072 tokens of context still available.
How this is chosen, since no benchmark is quoted: models are compared by size, a mixture-of-experts model counting at the geometric mean of its total and active parameters (a rule of thumb, not a measurement). The pick is the biggest class that fits with memory to spare and answers at 25 tokens per second or better, and within that class the most recent general-purpose release; coding and reasoning specialists are listed above but not recommended by default.
gpt-oss 20B is the largest model the card will hold, at 13.4 GB, with room for 81,920 tokens of context. It is a mixture-of-experts model reading only 3.6B parameters per token, so it is the faster of the two at an estimated 107 tokens per second; what it gives up is the depth of a dense model that reads all of its weights for every token.
It also fits at Q8_0, in about 14.2 GB, which is worth taking whenever the memory allows: Q8 is near-lossless where Q4 costs a little accuracy.
How much context actually fits
Model size is the question people ask; context is the one that bites. The KV cache grows linearly with the conversation, so a model that loads comfortably can still run out of memory halfway through a long document. On 16 GB, the context you get is whatever the weights leave behind, which is why two cards that run the same model can feel completely different in use. These are the largest contexts this card holds, at Q4_K_M weights.
| Model | Max context, FP16 KV | With Q8 KV cache |
|---|---|---|
| Gemma 4 12B | 131,072 tokens | 131,072 tokens |
| Qwen3 14B | 34,816 tokens | 40,960 tokens · model limit |
| Phi-4 14B | 16,384 tokens · model limit | 16,384 tokens · model limit |
| Ministral 3 14B | 37,888 tokens | 75,776 tokens |
Quantising the KV cache to Q8 roughly doubles what you can hold, at a quality cost most people never notice: Ministral 3 14B on this card goes from 37,888 to 75,776 tokens. Capped at 128k tokens, or at the model's own native window where that is smaller, marked "model limit": free memory beyond that point buys nothing. Rows that hold far more context than their size suggests are hybrid, sliding-window or latent-attention models, which cache only a fraction of what a standard transformer does; each model page shows the working.
Where this card stops
The first model out of reach is Devstral Small 2 24B: about 16.3 GB at Q4_K_M and 8k context, against 16 GB of usable memory. Dropping to Q3_K_M would need about 13.6 GB, which fits, though Q3 loses enough quality that a smaller model at Q4 is usually the better trade. Offloading the remainder to system RAM works and is ten to fifty times slower; it is a way to see a model run, not a way to use one.
Buy or rent
Worth it if 16 GB is what you need and a low-power card matters. If speed matters, an RTX 5060 Ti 16 GB holds the same and reads it 1.4 times as fast, and an RX 9070 XT reads it twice as fast. For the occasional larger model, renting by the hour costs less than buying up a tier.
Retail prices drift, and a figure written today would be wrong within a month, so this page does not carry one. Put the figure you are actually looking at, with your electricity rate and your real monthly hours, into the build vs rent calculator, which already knows this card draws 160 W under load. It returns the month owning becomes cheaper, or tells you it never does. Hours per month decides it far more often than the price does.
The honest take
AMD's 16 GB budget card: the memory is right, the bandwidth is low.
It holds what any 16 GB card holds, at 160 W, and that is the reason to buy it for models. It reads that memory at 320 GB/s, half an RX 9070 XT and below an RTX 3060's 360, so the larger models that fit are only just comfortable for chat. RDNA 4 brings FP8 and a place on AMD's ROCm 10.1 matrix for Linux and Windows 11; CUDA-only tools still do not run.
What it is good at
- 16 GB on a 160 W board that fits ordinary cases and power supplies.
- RDNA 4, so FP8 runs natively, unlike on the RX 7000 series.
- Listed on AMD's ROCm 10.1 matrix for Linux and Windows 11, and on Ollama's ROCm list for Linux.
What it is not
- 320 GB/s: half an RX 9070 XT, and less than an RTX 3060 12 GB at 360 GB/s. The larger the model, the more it shows.
- An 8 GB card carries the same name. Check the memory on the box.
- No CUDA: fine-tuning and much image and video tooling assume it.
The thing people get wrong: There are two RX 9060 XT cards, 8 GB and 16 GB, and AMD prints the same 320 GB/s for both (the 8 GB board is rated at 150 W). For language models only the 16 GB one is worth buying: the same speed with half the room.
Compared with the alternatives
- RTX 5060 Ti: 16 GB · 448 GB/s · best fit Gemma 4 12B at ~45 tok/s. The newest 16 GB GeForce below the high end, and a clear step up from the 4060 Ti it replaces.
- RX 9070 XT: 16 GB · 640 GB/s · best fit Gemma 4 12B at ~64 tok/s. AMD's RDNA 4 gaming card: 16 GB at 640 GB/s, with ROCm on Linux and Windows, and no CUDA.
- RTX 4060 Ti: 16 GB · 288 GB/s · best fit Gemma 4 12B at ~29 tok/s. More memory than its neighbours, and less bandwidth than a card two years older.
All 39 cards are compared side by side on the GPU index, and the same numbers from the model's point of view are on each model fit table. To check one specific pairing rather than read a table, use the can-I-run-it checker.
Questions
- Can an RX 9060 XT run a 70B model?
- Not at Q4_K_M in one card. Llama 3.3 70B needs about 45.8 GB at 8k context and this card can address 16 GB. Your options are a smaller model, a harsher quantisation with very little context, splitting across two cards, or renting a larger one for the hours you need it.
- What is the best model to run on an RX 9060 XT?
- Gemma 4 12B. At Q4_K_M and 8k context it needs about 8.2 GB of the 16 GB available and generates an estimated 32 tokens per second, which is comfortable for chat. It also fits at Q8, at about 14.2 GB, which is worth taking when it fits.
- How many tokens per second does an RX 9060 XT generate?
- It depends entirely on the size of the model, because generating a token means reading every active weight out of memory. At 320 GB/s and 70% efficiency, this card produces an estimated 48 tokens per second on an 8B model at Q4, and about 107 on the largest model it holds, gpt-oss 20B. Reading speed is roughly 5 to 8 words per second, so anything above 25 feels immediate. Treat 48 as a ceiling on 320 GB/s rather than a measurement; the higher a figure, the further real runtimes fall below it.
- Is 16 GB enough for running LLMs locally?
- It runs 12 of the 34 models tracked here at Q4_K_M with 8k of context, up to 21B parameters. The honest test is not the model list but the context: Gemma 4 12B on this card holds about 131,072 tokens before memory runs out.
- RX 9060 XT or RTX 5060 Ti for local models?
- Both address about the same memory, so they run the same models. On speed, the RTX 5060 Ti is faster: 448 GB/s against 320 GB/s, and bandwidth is what sets chat speed.
- RX 9060 XT 16 GB or RTX 5060 Ti 16 GB for AI?
- Both hold 16 GB, so the same models fit. The 5060 Ti reads memory at 448 GB/s against 320 GB/s, 1.4 times as fast, and runs CUDA, which most fine-tuning and image tools assume. The 9060 XT draws 160 W against 180 W and runs through ROCm or Vulkan. For language models alone, the 5060 Ti is the faster card.
- Does the RX 9060 XT support ROCm?
- Yes. AMD's ROCm 10.1 compatibility matrix lists it (as gfx1200) for Ubuntu, RHEL and Windows 11, and Ollama lists it for ROCm on Linux. On Windows, Ollama runs it through its Vulkan backend instead, which is on by default.
- Memory: weights (parameters × bytes per parameter) + KV cache (2 × layers × KV heads × head dim × context × bytes for a standard transformer; sliding-window, hybrid and latent-attention models counted as they actually cache) + 0.5 GB runtime + 4% of weights. Full derivation in the VRAM calculator.
- Speed: 70% × bandwidth ÷ active weight bytes, single stream, no batching, prompt processing excluded. A ceiling, not a measurement: the 2 of 12 figures marked ≤ are where real runtimes fall furthest below the number, because so few weights are read per token. See the tokens-per-second estimator.
- Card specification (manufacturer figures): 16 GB, 320 GB/s, 160 W; see AMD's own page for this card.
- Model architecture values come from each model's published config; the self-hosted LLM guide explains what each one changes.