Mac mini M4 64GBGünstige Wahl
64 GB Unified (+28)
C
Adds memory headroom for longer context windows and future model growth.8.9 tok/s Dekodierung
Adds memory headroom for longer context windows and future model growth.
ca. $1,099 MSRP
Can cognitivecomputations Dolphin3.0 R1 Mistral 24B run on MacBook Pro M3 Pro 36GB?
YES — Tight Fit
C46Usable
Estimated from fit model
cognitivecomputations Dolphin3.0 R1 Mistral 24B needs ~22.2 GB VRAM. MacBook Pro M3 Pro 36GB has 25.9 GB. With Q4_K_M quantization, expect ~8 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) — 22.2 GB, 7.5 tok/s, Tight fit
22.2 GB required25.9 GB available
86% VRAM used
Fit status
Tight fit
Decode
7.5 tok/s
TTFT
25884 ms
Safe context
37K
Memory
22.2 GB / 25.9 GB
Memory breakdown
Weights14.6 GB
KV Cache2.8 GB
Runtime0.9 GB
Headroom3.9 GB
See how fast it feels
See how fast it feelscognitivecomputations Dolphin3.0 R1 Mistral 24B on MacBook Pro M3 Pro 36GB
1st promptCold start — includes initialization
>Solve: If a train travels at 60 mph for 2.5 hours, how far does it go?
Let me work through this step by step.
Given information:
- Speed = 60 mph (miles per hour)
- Time = 2.5 hours
Formula: Distance = Speed x Time
Calculation:
Distance = 60 mph x 2.5 h
= 150 miles
The train travels 150 miles.
Double-check: at 60 mph the train covers 60 miles each hour. In half an hour it covers 30 miles. So in 2.5 hours: 60 + 60 + 30 = 150 miles. Confirmed.
2nd promptWarm — model ready, no init overhead
>What if the train stops for 20 minutes halfway?
The stop doesn't change the total distance covered while moving, but it affects total trip time.
Breakdown:
- First half: 1.25 hours at 60 mph = 75 miles
- Stop: 20 minutes (no distance)
- Second half: 1.25 hours at 60 mph = 75 miles
Total distance: 150 miles (same)
Total time: 2.5 h + 0.33 h = 2.83 hours
Average speed: 150 / 2.83 = 53 mph
The distance stays the same but average speed drops to 53 mph because of the stop.
Estimated: 7.5 tok/s decode · 25.9s TTFT (warm) · 19 tok/s prefill
What limits this setup
The model fits in shared memory, but shared-memory bandwidth is now the real limiter.
Fit does not mean dedicated-VRAM speed
Unified or shared memory can make a model technically fit, but sustained tokens per second may still trail a discrete high-bandwidth GPU with less total memory.
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
Prioritize bandwidth, not only capacity
If this workload feels slow, the next useful step is often a GPU tier with materially faster memory bandwidth rather than only a small bump in capacity.
Performance by workload
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Chat | C | Runs well | 7.5 tok/s | 14119 ms | 37K |
| Coding | C | Tight fit | 7.5 tok/s | 25884 ms | 37K |
| Agentic Coding | C | Runs with offload | 7.5 tok/s | 37650 ms | 37K |
| Reasoning | C | Tight fit | 7.5 tok/s | 30590 ms | 37K |
| RAG | C | Runs with offload | 7.5 tok/s | 47062 ms | 37K |
Quantization options
How cognitivecomputations Dolphin3.0 R1 Mistral 24B (24B params) fits at each quantization level on MacBook Pro M3 Pro 36GB (25.9 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 9.4 GB | Low | C48 |
Q3_K_S | 3 | 11.8 GB | Low | C49 |
NVFP4 | 4 | 13.4 GB | Medium | C50 |
Q4_K_M | 4 | 14.6 GB | Medium | C50 |
Q5_K_M | 5 | 17.3 GB | High | C49 |
Q6_KBest for your GPU | 6 | 19.7 GB | High | C49 |
Q8_0 | 8 | 25.7 GB | Very High | F0 |
F16 | 16 | 49.2 GB | Maximum | F0 |
Get started
Copy-paste commands to run cognitivecomputations Dolphin3.0 R1 Mistral 24B on your machine.
Run
lms load hf-bartowski--cognitivecomputations-dolphin3-0-r1-mistral-24b-gguf && lms server startUpgrade-Optionen
Hardware, die cognitivecomputations Dolphin3.0 R1 Mistral 24B gut ausführt
MacBook Pro M4 Pro 64GBBestes Preis-Leistungs-Verhältnis
64 GB Unified (+28)273 GB/s (+123)
C
Raises estimated decode speed by about 187%.21.5 tok/s Dekodierung
Raises estimated decode speed by about 187%.
Adds memory headroom for longer context windows and future model growth.
ca. $1,599 MSRP
MacBook Pro M4 Max 48GBApple-Upgrade
48 GB Unified (+12)546 GB/s (+396)
C
Raises estimated decode speed by about 356%.34.2 tok/s Dekodierung
Raises estimated decode speed by about 356%.
Adds memory headroom for longer context windows and future model growth.
ca. $2,499 MSRP
Frequently asked questions
Can MacBook Pro M3 Pro 36GB run cognitivecomputations Dolphin3.0 R1 Mistral 24B?
Yes, MacBook Pro M3 Pro 36GB can run cognitivecomputations Dolphin3.0 R1 Mistral 24B with a C grade (Tight fit). Expected decode speed: 7.5 tok/s.
How much VRAM does cognitivecomputations Dolphin3.0 R1 Mistral 24B need?
cognitivecomputations Dolphin3.0 R1 Mistral 24B (24B parameters) requires approximately 22.2 GB of memory with Q4_K_M quantization.
What is the best quantization for cognitivecomputations Dolphin3.0 R1 Mistral 24B?
The recommended quantization for cognitivecomputations Dolphin3.0 R1 Mistral 24B is Q4_K_M, which balances quality and memory efficiency.
What speed will cognitivecomputations Dolphin3.0 R1 Mistral 24B run at on MacBook Pro M3 Pro 36GB?
On MacBook Pro M3 Pro 36GB, cognitivecomputations Dolphin3.0 R1 Mistral 24B achieves approximately 7.5 tokens per second decode speed with a time-to-first-token of 25884ms using Q4_K_M quantization.
Can MacBook Pro M3 Pro 36GB run cognitivecomputations Dolphin3.0 R1 Mistral 24B for coding?
For coding workloads, cognitivecomputations Dolphin3.0 R1 Mistral 24B on MacBook Pro M3 Pro 36GB receives a C grade with 7.5 tok/s and 37K context.
What context window can cognitivecomputations Dolphin3.0 R1 Mistral 24B use on MacBook Pro M3 Pro 36GB?
On MacBook Pro M3 Pro 36GB, cognitivecomputations Dolphin3.0 R1 Mistral 24B can safely use up to 37K tokens of context. The model's official context limit is —, but available memory constrains the safe maximum.
What should I upgrade first if cognitivecomputations Dolphin3.0 R1 Mistral 24B feels slow on MacBook Pro M3 Pro 36GB?
Prioritize bandwidth, not only capacity. If this workload feels slow, the next useful step is often a GPU tier with materially faster memory bandwidth rather than only a small bump in capacity.
Is unified memory on MacBook Pro M3 Pro 36GB as fast as VRAM for cognitivecomputations Dolphin3.0 R1 Mistral 24B?
Not always. MacBook Pro M3 Pro 36GB 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/hf-bartowski--cognitivecomputations-dolphin3-0-r1-mistral-24b-gguf-on-m3-pro-36gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>
Preview: