Can StarCoder2 15B run on Mac Studio M3 Ultra 256GB?

YES — Runs Great

C46Usable

○Estimated from fit model

StarCoder2 15B needs ~40.6 GB VRAM. Mac Studio M3 Ultra 256GB has 184.3 GB. With Q5_K_M quantization, expect ~57 tok/s.

Runtime: llama.cppCapacity: RoomyBandwidth: HighStack: StandardBottleneck: Balanced

Operating mode

Choose the run profile you care about

Interactive favors responsiveness, while light API and scale-out lean harder on serving readiness. The fit stays the same, but the recommendation lens changes.

Current mode

Balanced

Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.

Capabilities:

Select quantization to explore

Q5_K_M (High quality) — 40.6 GB, 57.4 tok/s, Runs well

40.6 GB required184.3 GB available

22% VRAM used

Fit status

Runs well

Decode

57.4 tok/s

TTFT

3372 ms

Safe context

16K

Memory

40.6 GB / 184.3 GB

Memory breakdown

Weights10.8 GB

KV Cache1.2 GB

Runtime0.9 GB

Headroom27.6 GB

See how fast it feels

See how fast it feelsStarCoder2 15B on Mac Studio M3 Ultra 256GB
1st promptCold start — includes initialization
>Write a Python function to sort a list
Here's a Python function that sorts a list using quicksort:

def quicksort(arr):
 if len(arr) <= 1:
 return arr
 pivot = arr[len(arr) // 2]
 left = [x for x in arr if x < pivot]
 middle = [x for x in arr if x == pivot]
 right = [x for x in arr if x > pivot]
 return quicksort(left) + middle + quicksort(right)

# Example usage
numbers = [3, 6, 8, 10, 1, 2, 1]
print(quicksort(numbers)) # [1, 1, 2, 3, 6, 8, 10]

The algorithm picks a pivot, partitions the list into smaller, equal, and larger elements, then recursively sorts each partition and concatenates the results.
2nd promptWarm — model ready, no init overhead
>Now add type hints and a docstring
def quicksort(arr: list[int]) -> list[int]:
 """Sort a list using the quicksort algorithm.

 Args:
 arr: List of integers to sort.

 Returns:
 A new sorted list.
 """
 if len(arr) <= 1:
 return arr
 pivot = arr[len(arr) // 2]
 left = [x for x in arr if x < pivot]
 middle = [x for x in arr if x == pivot]
 right = [x for x in arr if x > pivot]
 return quicksort(left) + middle + quicksort(right)

Added type hints for the input and return type, plus a Google-style docstring.
Estimated: 57.4 tok/s decode · 3.4s TTFT (warm) · 144 tok/s prefill

What limits this setup

This setup is broadly balanced for this model.

Shared-memory contention still exists

The OS, browser, and inference runtime all compete for the same physical memory pool, so real-world headroom is less forgiving than raw capacity suggests.

Best improvement path

Performance by workload

Workload	Grade	Fit	Decode	TTFT	Context
Chat	C	Runs well	57.4 tok/s	1839 ms	16K
Coding	C	Runs well	57.4 tok/s	3372 ms	16K
Agentic Coding	C	Runs well	57.4 tok/s	4904 ms	16K
Reasoning	C	Runs well	57.4 tok/s	3985 ms	16K
RAG	C	Runs well	57.4 tok/s	6130 ms	16K

Quantization options

How StarCoder2 15B (15B params) fits at each quantization level on Mac Studio M3 Ultra 256GB (184.3 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	5.9 GB	Low	D38
Q3_K_S	3	7.4 GB	Low	D38
NVFP4	4	8.4 GB	Medium	D38
Q4_K_M	4	9.2 GB	Medium	D38
Q5_K_M	5	10.8 GB	High	D39
Q6_K	6	12.3 GB	High	D39
Q8_0	8	16.1 GB	Very High	D39
F16Best for your GPU	16	30.7 GB	Maximum	C40

Get started

Copy-paste commands to run StarCoder2 15B on your machine.

Run

docker run --rm -it ghcr.io/ggerganov/llama.cpp:full \
 --hf-repo "bigcode/starcoder2-15b" \
 --hf-file "starcoder2-15b-Q5_K_M.gguf" \
 -c 4096 -ngl 99

Upgrade options