| Model | Verdict | Decode tok/s | Max ctx | Size | Chat | Coding | |
|---|---|---|---|---|---|---|---|
| Qwen/Qwen3-30B-A3B-Instruct-2507 lmstudio-community/Qwen3-30B-A3B-Instruct-2507-MLX-8bit · MLX-8BIT | Fits | 42.116.8–46.3 | 256K | 30.5B | p85 | – | |
| google/gemma-4-31B-it lmstudio-community/gemma-4-31B-it-MLX-8bit · MLX-8BIT | Fits | 9.103.60–10.0 | 256K | 31.3B | p63 | – | |
| Qwen/Qwen3.5-27B unsloth/Qwen3.5-27B-GGUF · Q8_0 | Fits | 11.54.60–12.7 | 256K | 27.8B | p60 | p46 | |
| Qwen/Qwen3.8-27B lmstudio-community/Qwen3.8-27B-MLX-8bit · MLX-8BIT | Fits | 11.24.50–12.3 | 256K | 27.8B | p59 | p33 | |
| Qwen/Qwen3.6-27B lmstudio-community/Qwen3.6-27B-MLX-8bit · MLX-8BIT | Fits | 11.24.50–12.3 | 256K | 27.8B | p51 | p54 | |
| Qwen/Qwen3.5-35B-A3B unsloth/Qwen3.5-35B-A3B-GGUF · Q8_0 | Fits | 66.226.5–72.8 | 256K | 36B | p49 | p35 | |
| google/gemma-4-26B-A4B-it lmstudio-community/gemma-4-26B-A4B-it-MLX-8bit · MLX-8BIT | Fits | 53.721.5–59.1 | 256K | 25.8B | p49 | – | |
| Qwen/Qwen3.5-9B lmstudio-community/Qwen3.5-9B-MLX-8bit · MLX-8BIT | Fits | 30.212.1–33.2 | 256K | 9.7B | p46 | – | |
| ornith-ai/Ornith-1.5-9B AtomicChat/Ornith-1.5-9B-GGUF · Q8_0 | Fits | 32.412.9–35.6 | 256K | 9.7B | p46 | p24 | |
| Qwen/Qwen3.6-35B-A3B lmstudio-community/Qwen3.6-35B-A3B-MLX-8bit · MLX-8BIT | Fits | 65.526.2–72.0 | 256K | 36B | p43 | p41 | |
| openbmb/MiniCPM-V-4.7-35B-A3B openbmb/MiniCPM-V-4.7-35B-A3B-gguf · Q8_0 | Fits | 66.326.5–72.9 | 256K | 35.2B | p42 | – | |
| ornith-ai/Ornith-1.0-35B lmstudio-community/Ornith-1.0-35B-MLX-8bit · MLX-8BIT | Fits | 65.626.3–72.2 | 256K | 35.1B | – | p40 | |
| openai/gpt-oss-20b MXFP4 | Fits | 49.519.8–54.5 | 128K | 21.1B | p39 | p15 | |
| Qwen/Qwen3.5-4B unsloth/Qwen3.5-4B-GGUF · Q8_0 | Fits | 51.120.4–56.2 | 256K | 4.7B | p39 | – | |
| google/gemma-4-12B-it unsloth/gemma-4-12b-it-GGUF · Q8_0 | Fits | 20.117.1–23.1 | 256K | 12B | p38 | – | |
| Qwen/Qwen3-4B-Instruct-2507 BF16 | Fits | 27.723.6–31.9 | 256K | 4B | p38 | – | |
| nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 BF16 | Fits | 67.126.8–73.8 | 256K | 32.9B | p37 | p9 | |
| nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4 MIXED_PRECISION | Fits | 81.032.4–89.1 | 1M | 47.1B | p36 | p9 | |
| nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 BF16 | Fits | 68.727.5–75.6 | 256K | 31.6B | p33 | – | |
| poolside/Laguna-XS-2.1 AtomicChat/Laguna-XS-2.1-MLX-8bit · MLX-8BIT | Fits | 72.128.8–79.3 | 256K | 33.4B | – | p30 | |
| poolside/Laguna-XS.2 poolside-laguna-hackathon/Laguna-XS.2-GGUF · Q4_K_M | Fits | 52.921.2–58.2 | 256K | 33.4B | – | p25 | |
| ornith-ai/Ornith-1.0-9B lmstudio-community/Ornith-1.0-9B-MLX-8bit · MLX-8BIT | Fits | 30.212.1–33.3 | 256K | 9.4B | – | p20 | |
| ornith-ai/Ornith-1.5-35B-A3B-NVFP4 MIXED_PRECISION | Fits | 80.432.1–88.4 | 256K | 34B | – | – | |
| nvidia/Qwen3.6-35B-A3B-NVFP4 MIXED_PRECISION | Fits | 80.432.1–88.4 | 256K | 34B | – | – | |
| Qwen/Qwen3-Coder-30B-A3B-Instruct-FP8 FP8 | Fits | 42.216.9–46.5 | 256K | 30.5B | – | – | |
| prism-ml/Bonsai-27B-mlx-1bit MLX-1BIT | Fits | 41.416.5–45.5 | 256K | 26.4B | – | – | |
| ornith-ai/Ornith-1.5-9B-GGUF Q8_0 | Fits | 31.812.7–35.0 | 256K | 9.2B | – | – | |
| Qwen/Qwen3-30B-A3B BF16 | Fits | 31.212.5–34.4 | 40K | 30.5B | – | – | |
| Qwen/Qwen3-Coder-30B-A3B-Instruct BF16 | Fits | 31.212.5–34.4 | 153K | 30.5B | – | – | |
| prism-ml/Ternary-Bonsai-27B-mlx-2bit MLX-2BIT | Fits | 30.212.1–33.2 | 256K | 26.4B | – | – | |
| Qwen/Qwen3-14B lmstudio-community/Qwen3-14B-MLX-8bit · MLX-8BIT | Fits | 17.67.00–19.4 | 40K | 14.8B | – | – | |
| Qwen/Qwen2.5-14B-Instruct Qwen/Qwen2.5-14B-Instruct-GGUF · Q8_0 | Fits | 17.014.4–19.5 | 32K | 14.8B | – | – | |
| Qwen/Qwen2.5-Coder-14B-Instruct Qwen/Qwen2.5-Coder-14B-Instruct-GGUF · Q4_K_M | Fits | 15.413.1–17.7 | 32K | 14.8B | – | – | |
| RadixArk/Qwen3.8-27B-NVFP4 MIXED_PRECISION | Fits | 14.55.80–16.0 | 256K | 26.4B | – | – | |
| farbodtavakkoli/OTel-2.0-LLM-31B-IT wfakhri/OTel-2.0-LLM-31B-IT-GGUF · Q8_0 | Fits | 9.307.90–10.7 | 256K | 31.3B | – | – | |
| Qwen/Qwen2.5-32B-Instruct Qwen/Qwen2.5-32B-Instruct-GGUF · Q8_0 | Fits | 8.607.30–9.90 | 32K | 32.8B | – | – | |
| Qwen/Qwen2.5-Coder-32B-Instruct lmstudio-community/Qwen2.5-Coder-32B-Instruct-MLX-8bit · MLX-8BIT | Fits | 8.603.40–9.40 | 32K | 32.8B | – | – | |
| Qwen/Qwen3-32B lmstudio-community/Qwen3-32B-MLX-8bit · MLX-8BIT | Fits | 8.403.30–9.20 | 40K | 32.8B | – | – |
Quality = benchmark percentile against all indexed models (p50 = median). Speeds are single-user decode ranges.