MacBook Air M4 24GBコスパ重視
24 GB Unified (+8)120 GB/s (+20)
C
Raises estimated decode speed by about 33%.21.9 tok/sデコード
Raises estimated decode speed by about 33%.
Adds memory headroom for longer context windows and future model growth.
〜$1,099 MSRP
Can Granite 3.1 8B run on MacBook Air M2 16GB?
YES — Tight Fit
C53Usable
Estimated from fit model
Granite 3.1 8B needs ~9.5 GB VRAM. MacBook Air M2 16GB has 11.5 GB. With Q4_K_M quantization, expect ~17 tok/s.
Runtime: llama.cppCapacity: TightBandwidth: Very lowStack: StandardBottleneck: Memory bandwidth
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) — 9.5 GB, 16.5 tok/s, Tight fit
9.5 GB required11.5 GB available
83% VRAM used
Fit status
Tight fit
Decode
16.5 tok/s
TTFT
11757 ms
Safe context
33K
Memory
9.5 GB / 11.5 GB
Memory breakdown
Weights4.9 GB
KV Cache2.0 GB
Runtime0.9 GB
Headroom1.7 GB
See how fast it feels
See how fast it feelsGranite 3.1 8B on MacBook Air M2 16GB
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: 16.5 tok/s decode · 11.8s TTFT (warm) · 41 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 | 16.5 tok/s | 6413 ms | 33K |
| Coding | C | Tight fit | 16.5 tok/s | 11757 ms | 33K |
| Agentic Coding | C | Runs with offload | 16.5 tok/s | 17101 ms | 33K |
| Reasoning | C | Tight fit | 16.5 tok/s | 13895 ms | 33K |
| RAG | C | Runs with offload | 16.5 tok/s | 21377 ms | 33K |
Inference speed
Granite 3.1 8B inference speed — tokens per second by GPU & Mac
Estimated decode speed (tokens/sec) for Granite 3.1 8B at Q4_K_M across popular GPUs and Apple Silicon, using the fastest local runtime per device. Fastest is RTX 5090 32GB at ~112 tok/s. Speed is memory-bandwidth bound, so cards that fit the whole model in VRAM run far faster than ones that offload to system RAM.
| GPU / Mac | Memory | Quant | Speed (tok/s) | Fits? |
|---|---|---|---|---|
 RTX 5090 32GB | 32 GB | Q4_K_M | 112.0 | Fits |
| RTX 4090 24GB | 24 GB | Q4_K_M | 112.0 | Fits |
| RTX 4080 Super 16GB | 16 GB | Q4_K_M | 112.0 | Fits |
| RTX 3090 24GB | 24 GB | Q4_K_M | 112.0 | Fits |
RX 7900 XTX 24GB | 24 GB | Q4_K_M | 112.0 | Fits |
Mac Studio M3 Ultra 256GB | 256 GB | Q4_K_M | 112.0 | Fits |
Mac Studio M2 Ultra 128GB | 128 GB | Q4_K_M | 112.0 | Fits |
Mac Studio M1 Ultra 128GB | 128 GB | Q4_K_M | 111.5 | Fits |
| RTX 4070 12GB | 12 GB | Q4_K_M | 95.8 | Fits |
MacBook Pro M4 Max 128GB | 128 GB | Q4_K_M | 87.1 | Fits |
MacBook Pro M4 Max 64GB | 64 GB | Q4_K_M | 87.1 | Fits |
MacBook Pro M3 Max 64GB | 64 GB | Q4_K_M | 60.8 | Fits |
| RTX 3060 12GB | 12 GB | Q4_K_M | 60.2 | Fits |
MacBook Pro M1 Max 64GB | 64 GB | Q4_K_M | 55.7 | Fits |
MacBook Pro M4 Pro 48GB | 48 GB | Q4_K_M | 53.3 | Fits |
| RTX 4060 8GB | 8 GB | Q4_K_M | 32.9 | Offloads |
Estimates for single-stream decoding at Q4_K_M; real tokens/sec varies with prompt length, context, batch size, and runtime build. Prompt processing (prefill) is faster than the decode figures shown here.
Quantization options
How Granite 3.1 8B (8B params) fits at each quantization level on MacBook Air M2 16GB (11.5 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 3.1 GB | Low | C54 |
Q3_K_S | 3 | 3.9 GB | Low | B56 |
NVFP4 | 4 | 4.5 GB | Medium | B56 |
Q4_K_M | 4 | 4.9 GB | Medium | B57 |
Q5_K_M | 5 | 5.8 GB | High | B57 |
Q6_KBest for your GPU | 6 | 6.6 GB | High | B57 |
Q8_0 | 8 | 8.6 GB | Very High | F0 |
F16 | 16 | 16.4 GB | Maximum | F0 |
Get started
Copy-paste commands to run Granite 3.1 8B on your machine.
Run
ollama run granite3.1-denseアップグレードオプション
Granite 3.1 8Bを快適に動かすハードウェア
MacBook Pro M3 24GBコスパ最良
24 GB Unified (+8)
C
Adds memory headroom for longer context windows and future model growth.17.2 tok/sデコード
Adds memory headroom for longer context windows and future model growth.
〜$1,099 MSRP
MacBook Air M3 24GBAppleアップグレード
24 GB Unified (+8)
C
Adds memory headroom for longer context windows and future model growth.17.2 tok/sデコード
Adds memory headroom for longer context windows and future model growth.
〜$1,099 MSRP
RTX 3080 Ti 12GB最大の飛躍912 GB/s (+812)
B
Raises estimated decode speed by about 482%.96 tok/sデコード
Raises estimated decode speed by about 482%.
〜$1,199 MSRP
Frequently asked questions
Can MacBook Air M2 16GB run Granite 3.1 8B?
Yes, MacBook Air M2 16GB can run Granite 3.1 8B with a C grade (Tight fit). Expected decode speed: 16.5 tok/s.
How much VRAM does Granite 3.1 8B need?
Granite 3.1 8B (8B parameters) requires approximately 9.5 GB of memory with Q4_K_M quantization.
What is the best quantization for Granite 3.1 8B?
The recommended quantization for Granite 3.1 8B is Q4_K_M, which balances quality and memory efficiency.
What speed will Granite 3.1 8B run at on MacBook Air M2 16GB?
On MacBook Air M2 16GB, Granite 3.1 8B achieves approximately 16.5 tokens per second decode speed with a time-to-first-token of 11757ms using Q4_K_M quantization.
Can MacBook Air M2 16GB run Granite 3.1 8B for coding?
For coding workloads, Granite 3.1 8B on MacBook Air M2 16GB receives a C grade with 16.5 tok/s and 33K context.
What context window can Granite 3.1 8B use on MacBook Air M2 16GB?
On MacBook Air M2 16GB, Granite 3.1 8B can safely use up to 33K tokens of context. The model's official context limit is 128K, but available memory constrains the safe maximum.
Is unified memory on MacBook Air M2 16GB as fast as VRAM for Granite 3.1 8B?
Not always. MacBook Air M2 16GB can often fit larger models thanks to unified memory, but a discrete GPU with dedicated high-bandwidth VRAM may still decode faster once the model fits. For this combination, the important distinction is capacity versus sustained throughput.
Embed this result▼
Paste this snippet into any page to show a live fit card.
<iframe src="https://willitrunai.com/embed/granite-3.1-8b-on-m2-air-16gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>
Preview: