What can you run with 16GB of RAM?

With 16GB RAM you can run 8 of the 30 open-source models we track — up to about 22B parameters at Q4 quantization. The most popular that fits is Qwen3.5-9B, needing 6.8 GB and generating around ~10 tok/s on typical hardware of this size.

Typical machines: MacBook Air M4 · Mid-range Windows laptops · Mac mini (base). After the operating system takes its share, about 14 GB is free for a model.

Having the memory is not the same as having the speed

16GB RAM decides whether a model loads. How fast it answers is decided by memory bandwidth — how quickly your machine can read the model, which it must do once for every word it writes. Machines with 16GB RAM span 51–137 GB/s, so the fastest is about 2.7× quicker than the slowest with identical capacity. The table quotes a DDR5-5600; pick yours below to see the difference. Why bandwidth and not the processor →

51 GB/sDesktop or older laptop — DDR4-3200· 3,200 MT/s × 128-bit dual channel ÷ 8 = 51.2 GB/s
68 GB/sThin laptop — LPDDR4X-4266· 4,266 MT/s × 128-bit ÷ 8 = 68.3 GB/s (soldered, 2 channels)
90 GB/sDesktop or laptop — DDR5-5600· 5,600 MT/s × 128-bit dual channel ÷ 8 = 89.6 GB/s
102 GB/sModern thin laptop — LPDDR5-6400· 6,400 MT/s × 128-bit ÷ 8 = 102.4 GB/s (soldered, 2 channels)
120 GB/sMacBook Air or Mac mini — M4· Apple published specification for the M4: 120 GB/s
137 GB/sCurrent thin laptop — LPDDR5X-8533· 8,533 MT/s × 128-bit ÷ 8 = 136.5 GB/s (soldered, 2 channels)

Models that run on 16GB RAM in 2026

ModelSizeMemoryEst. speedFeels like
Qwen3.5-9BAlibaba / Qwen9.7B6.8 GBRuns well~10 tok/sComfortable
gpt-oss-20bOpenAI · MoE22B13 GBTight~25 tok/s4.3B activeFast
Qwen3.5-4BAlibaba / Qwen4.7B3.8 GBRuns well~21 tok/sComfortable
Qwen3.5-0.8BAlibaba / Qwen873M1.3 GBRuns well~110 tok/sFast
Qwen3.5-2BAlibaba / Qwen2.3B2.0 GBRuns well~46 tok/sFast
Olmo-3-7B-InstructAi27.3B7.2 GBRuns well~13 tok/sComfortable
GLM-4.6V-FlashZhipu AI10B7.0 GBRuns well~9.5 tok/sComfortable
ERNIE-4.5-0.3B-PTBaidu AI361M853 MBRuns well~243 tok/sFast

All figures at Q4_K_M quantization, 4K context. “Memory” includes the model, its context cache and runtime overhead. Speed is calculated, not benchmarked: 90 GB/s × 0.65 ÷ bytes read per token, assuming a DDR5-5600. Scale it by your own machine's bandwidth from the list above. Mixture-of-experts models read only their active parameters per token, which is why some large ones outrun smaller dense models.

Image generators that also fit

These make pictures rather than text. Their memory use scales with the resolution you generate at, not conversation length, so treat these figures as a floor rather than a ceiling.

What you'd need more memory for

The next models up, and what they ask for.

See what 32GB gets you →

Other hardware

Verified against HuggingFace on 2026-08-01New to this? Start here →Runs on a laptop and up · last 12 months only