Best Local LLMs for 96GB RAM

Ranked open-weight models that run well on a 96GB machine — with estimated speed and the exact ollama command for each.

Try: RTX 4090, MacBook Pro M4, 16GB, iPhone 17 Pro

Best local models for ~96GB

12 picks

Estimates assume a representative Apple-Silicon machine with 96GB unified memory. Tok/s are ModelFit estimates, not measured benchmarks. Run the wizard for figures tuned to your exact chip.

01

Qwen3.6 35B-A3B

Qwen / 35B / Q4_K_M / ~22 GB

Best for: Reasoning, Coding, Agents·Perf: ~38.6 tok/s (est.) · first token ~1.5s

Perfect fit

Best for reasoning, coding, agents. Strong fit for 96 GB RAM with balanced speed and quality.

ollama
$ollama run qwen3.6:35b-a3b
02

Qwen3 30B

Qwen / 30B / Q4_K_M / ~22 GB

Best for: Quality, Coding·Perf: ~44.4 tok/s (est.) · first token ~1.5s

Perfect fit

Best for quality, coding. Strong fit for 96 GB RAM with balanced speed and quality.

ollama
$ollama run qwen3:30b
03

Gemma 3 27B Instruct

Gemma / 27B / Q4_K_M / ~21 GB

Best for: Quality, Coding·Perf: ~48.8 tok/s (est.) · first token ~0.7s

Perfect fit

Best for quality, coding. Strong fit for 96 GB RAM with balanced speed and quality.

ollama
$ollama run gemma3:27b
04

Llama 3.3 70B Instruct

Llama / 70B / Q4_K_M / ~42 GB

Best for: Quality, Coding·Perf: ~20.7 tok/s (est.) · first token ~1.7s

Runs well

Best for quality, coding. Strong fit for 96 GB RAM with balanced speed and quality.

ollama
$ollama run llama3.3:70b-instruct-q4_K_M
05

Mixtral 8x7B Instruct

Mistral / 46.7B / Q4_K_M / ~30 GB

Best for: Coding, Quality·Perf: ~29.8 tok/s (est.) · first token ~1.6s

Runs well

Best for coding, quality. Strong fit for 96 GB RAM with balanced speed and quality.

ollama
$ollama run mixtral:8x7b
06

Gemma 2 27B Instruct

Gemma / 27B / Q4_K_M / ~21 GB

Best for: Quality, Coding·Perf: ~48.8 tok/s (est.) · first token ~0.7s

Perfect fit

Best for quality, coding. Strong fit for 96 GB RAM with balanced speed and quality.

ollama
$ollama run gemma2:27b-instruct-q4_K_M
07

Llama 3.1 70B Instruct

Llama / 70B / Q4_K_M / ~42 GB

Best for: Quality, Coding·Perf: ~20.7 tok/s (est.) · first token ~1.7s

Runs well

Best for quality, coding. Strong fit for 96 GB RAM with balanced speed and quality.

ollama
$ollama run llama3.1:70b-instruct-q4_K_M
08

DeepSeek-R1 Distill Llama 70B

DeepSeek / 70B / Q4_K_M / ~42 GB

Best for: Reasoning, Quality·Perf: ~20.7 tok/s (est.) · first token ~1.7s

Runs well

Best for reasoning, quality. Strong fit for 96 GB RAM with balanced speed and quality.

ollama
$ollama run deepseek-r1:70b
09

Qwen3.5 35B-A3B Instruct

Qwen / 35B / Q4_K_M / ~20 GB

Best for: Reasoning, Coding, Agent scenarios·Perf: ~38.6 tok/s (est.) · first token ~1.5s

Perfect fit

Best for reasoning, coding, agent scenarios. Strong fit for 96 GB RAM with balanced speed and quality.

ollama
$ollama run qwen3.5:35b-a3b
10

Qwen3.5 27B Instruct

Qwen / 27B / Q4_K_M / ~16 GB

Best for: Chat, Coding, Complex reasoning·Perf: ~48.8 tok/s (est.) · first token ~0.7s

Perfect fit

Best for chat, coding, complex reasoning. Strong fit for 96 GB RAM with balanced speed and quality.

ollama
$ollama run qwen3.5:27b
11

Qwen3.6 27B

Qwen / 27B / Q4_K_M / ~18 GB

Best for: Coding, Quality, Long context·Perf: ~48.8 tok/s (est.) · first token ~0.7s

Perfect fit

Best for coding, quality, long context. Strong fit for 96 GB RAM with balanced speed and quality.

ollama
$ollama run qwen3.6:27b
12

Gemma 4 26B-A4B

Gemma / 26B / Q4_K_M / ~16 GB

Best for: Chat, Coding, Multimodal·Perf: ~50.5 tok/s (est.) · first token ~0.6s

Perfect fit

Best for chat, coding, multimodal. Strong fit for 96 GB RAM with balanced speed and quality.

ollama
$ollama run gemma4:26b

Browse by RAM tier

8 GB16 GB24 GB32 GB48 GB64 GB96 GB128 GB

Want an exact recommendation?

The wizard tunes picks and speed estimates to your exact device, chip, and RAM.

Open ModelFit Wizard Browse all devices