Yes — with 7.2 GB to spare
Qwen3.5 4B at Q4_K_M fits your M4 · 16 GB entirely in unified memory at 8K context, at an estimated 26 tokens per second. Past 239K the KV cache pushes it over — quantise the cache to q8_0, or step down a quantisation, to go longer.
The 8 GB coding agent. Q4 lands near 3.4 GB, leaving room for a real context window.
Hardware requirements What hardware do I need to run Qwen3.5 4B? →At Q4_K_M with an 8K context: a 6 GB GPU (RTX 3060 Laptop) or a 8 GB Mac (M3). Minimum, recommended, high-quality and long-context tiers, with all 114 devices checked at every quantisation.
The VRAM budget
Quantisation ladder
| Quant | Weights | Total @ 8K | Max context | Tok/s | Quality | Fit |
|---|---|---|---|---|---|---|
| F16 | 8.7 GB | 9.5 GB | 45K | 7.7 | Reference | Long context |
| Q8_0 | 4.6 GB | 5.5 GB | 175K | 15 | −0.1% ppl | Long context |
| Q6_K | 3.6 GB | 4.4 GB | 209K | 19 | −0.4% ppl | Long context |
| Q5_K_M | 3.1 GB | 3.9 GB | 224K | 22 | −0.8% ppl | Long context |
| Q4_K_M | 2.6 GB | 3.5 GB | 239K | 26 | −1.9% ppl | Recommended |
| Q3_K_M | 2.1 GB | 3.0 GB | 255K | 32 | −5.4% ppl | Long context |
| Q2_K | 1.8 GB | 2.7 GB | 256K | 37 | −15% ppl | Long context |
Quality is the published perplexity delta against f16 weights. Max context assumes an f16 KV cache; q8_0 roughly doubles it. Only 8 of its 32 blocks keep a per-token KV cache; the rest are linear-attention, Mamba or convolution blocks with a fixed-size state.
How to run it
$ pip install mlx-lm $ mlx_lm.generate --model mlx-community/Qwen3.5-4B-4bit \ --max-tokens 512 --prompt "Hello"
Apple's own array framework. The fastest path on Apple Silicon. More on MLX.
Related models on an M4 · 16 GB
| Model | Total | Tok/s est. | |
|---|---|---|---|
| Qwen3.5 2B 2.27B · same family |
Q4_K_M · 2.0 GB | 53 | Runs great |
| Qwen3.5 9B 9.65B · same family |
Q4_K_M · 6.3 GB | 12 | Runs great |
| Qwen3.5 0.8B 0.87B · same family |
Q4_K_M · 1.2 GB | 137 | Runs great |
| Qwen3.5 122B-A10B 125B MoE · same family |
Q4_K_M · 71.1 GB | — | Won't fit |
| Gemma 3 4B 4.3B · similar size |
Q4_K_M · 3.3 GB | 28 | Runs great |
| Gemma 4 E2B 5.1B · similar size |
Q4_K_M · 3.5 GB | 23 | Runs great |
| Qwen3 4B 4.02B · similar size |
Q4_K_M · 4.0 GB | 30 | Runs great |
Qwen3.5 4B on other hardware
| Device | Total | Tok/s est. | |
|---|---|---|---|
| GeForce RTX 3060 Laptop 5.3 GB usable · 336 GB/s |
3.5 GB | 78 | Runs great |
| GeForce RTX 4060 Laptop 7.0 GB usable · 256 GB/s |
3.5 GB | 59 | Runs great |
| M3 · 8 GB 5.4 GB usable · 100 GB/s |
3.5 GB | 21 | Runs great |
| GeForce RTX 3060 12 GB 10.6 GB usable · 360 GB/s |
3.5 GB | 83 | Runs great |
| GeForce RTX 4090 22.4 GB usable · 1008 GB/s |
3.5 GB | 233 | Runs great |
| Radeon RX 7900 XTX 22.4 GB usable · 960 GB/s |
3.5 GB | 222 | Runs great |
| M3 Max · 64 GB 48.0 GB usable · 400 GB/s |
3.5 GB | 85 | Runs great |