1344 GB/s (+544)
B
Raises estimated decode speed by about 131%.85.9 tok/s decode
Raises estimated decode speed by about 131%.
~$4,999 MSRP
Can Codestral 22B run on Mac Studio M2 Ultra 64GB?
YES — Runs Great
B60Good
○Estimated from fit model
Codestral 22B needs ~23.7 GB VRAM. Mac Studio M2 Ultra 64GB has 46.1 GB. With Q4_K_M quantization, expect ~35 tok/s.
Runtime: llama.cppCapacity: RoomyBandwidth: HighStack: StandardBottleneck: Balanced
Share:
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.
Select quantization to explore
Q4_K_M (Medium quality) — 23.7 GB, 37.2 tok/s, Runs well
23.7 GB required46.1 GB available
51% VRAM used
Fit status
Runs well
Decode
37.2 tok/s
TTFT
5209 ms
Safe context
33K
Memory
23.7 GB / 46.1 GB
Memory breakdown
Weights13.4 GB
KV Cache2.4 GB
Runtime0.9 GB
Headroom6.9 GB
See how fast it feels
See how fast it feelsCodestral 22B on Mac Studio M2 Ultra 64GB
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: 37.2 tok/s decode · 5.2s TTFT (warm) · 93 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 | B | Runs well | 34.6 tok/s | 3054 ms | 33K |
| Coding | B | Runs well | 34.6 tok/s | 5599 ms | 33K |
| Agentic Coding | B | Runs well | 34.6 tok/s | 8145 ms | 33K |
| Reasoning | B | Runs well | 34.6 tok/s | 6617 ms | 33K |
| RAG | B | Runs well | 34.6 tok/s | 10181 ms | 33K |
Quantization options
How Codestral 22B (22B params) fits at each quantization level on Mac Studio M2 Ultra 64GB (46.1 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 8.6 GB | Low | C53 |
Q3_K_S | 3 | 10.8 GB | Low | C53 |
NVFP4 | 4 |
Get started
Copy-paste commands to run Codestral 22B on your machine.
Run
ollama run codestralUpgrade options