localmaxxing
Community leaderboard for local LLM inference speed across model, hardware, engine, quantization, context length, and batch-size configurations.
543rows
output_throughputprimary metric
2026-05-06sampled
Metadata
Metrics
Output Throughput, Time to First Token (lower is better), Total Throughput, Prefill Throughput, Peak VRAM (lower is better), Context Length, Batch Size
| Rank | Subject | Output Throughput | Model Match | Provenance | Sampled |
|---|---|---|---|---|---|
| 1 | Qwen3.5-0.8B-Base on NVIDIA H200 NVL (vllm BF16) | 2665.14 | — | Imported | 2026-05-06 |
| 2 | gpt-oss-20b on 2x AMD Radeon AI Pro R9700 (vllm MXFP4_MOE) | 991.10 | — | Imported | 2026-05-06 |
| 3 | Qwen3.5-122B-A10B-FP8 on 4x NVIDIA H200 SXM (vllm FP8) | 878.38 | — | Imported | 2026-05-06 |
| 4 | Qwen3.5-9B on RX 7900 XTX (hipfire MQ4) | 575.25 | — | Imported | 2026-05-06 |
| 5 | Qwen3.6-35B-A3B-NVFP4 on NVIDIA RTX PRO 6000 Blackwell (vllm NVFP4) | 506.20 | — | Imported | 2026-05-06 |
| 6 | MiniMax-M2.5 on 4x NVIDIA H200 SXM (vllm FP8) | 503.74 | — | Imported | 2026-05-06 |
| 7 | MiniMax-M2.1 on 4x NVIDIA H200 SXM (vllm FP8) | 499.20 | — | Imported | 2026-05-06 |
| 8 | MiniMax-M2.7 on 4x NVIDIA H200 SXM (vllm FP8) | 495.88 | — | Imported | 2026-05-06 |
| 9 | MiniMax-M2 on 4x NVIDIA H200 SXM (vllm FP8) | 492.60 | — | Imported | 2026-05-06 |
| 10 | MiniMax-M2.7 on 2x NVIDIA H200 NVL (vllm FP8) | 338.89 | — | Imported | 2026-05-06 |
| 11 | Qwen3.5-9B on RX 7900 XTX (hipfire MQ4) | 336.98 | — | Imported | 2026-05-06 |
| 12 | MiniMax-M2.5 on 2x NVIDIA H200 NVL (vllm FP8) | 333.93 | — | Imported | 2026-05-06 |
| 13 | MiniMax-M2.1 on 2x NVIDIA H200 NVL (vllm FP8) | 332.90 | — | Imported | 2026-05-06 |
| 14 | Qwen3.5-9B on RX 7900 XTX (hipfire MQ4) | 322.60 | — | Imported | 2026-05-06 |
| 15 | MiniMax-M2 on 2x NVIDIA H200 NVL (vllm FP8) | 302.51 | — | Imported | 2026-05-06 |
| 16 | NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 on NVIDIA GeForce RTX 5090 (ollama Q4_K_M) | 285.60 | — | Imported | 2026-05-06 |
| 17 | NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4 on NVIDIA H200 NVL (vLLM NVFP4) | 261.57 | — | Imported | 2026-05-06 |
| 18 | Qwen3.6-35B-A3B-FP8 on NVIDIA RTX PRO 6000 Blackwell (vllm FP8) | 253.70 | — | Imported | 2026-05-06 |
| 19 | Qwen3.5-27B on RX 7900 XTX (hipfire MQ4) | 250.25 | — | Imported | 2026-05-06 |
| 20 | GLM-5-NVFP4 on 4x NVIDIA H200 SXM (vllm NVFP4) | 217.23 | — | Imported | 2026-05-06 |
| 21 | Qwen3.6-27B-DFlash on RTX 5090 (llama.cpp Q4_K_M) | 215.43 | — | Imported | 2026-05-06 |
| 22 | Qwen3.5-27B on RX 7900 XTX (hipfire MQ4) | 201.11 | — | Imported | 2026-05-06 |
| 23 | GLM-5.1-NVFP4-MTP on 4x NVIDIA H200 SXM (vllm NVFP4) | 197.03 | — | Imported | 2026-05-06 |
| 24 | Qwen3.6-35B-A3B on RTX 5090 (llama.cpp UD-Q4_K_XL) | 190 | — | Imported | 2026-05-06 |
| 25 | NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4 on NVIDIA RTX PRO 6000 Blackwell (vLLM NVFP4) | 182.90 | — | Imported | 2026-05-06 |
| 26 | Qwen3.5-27B on RX 7900 XTX (hipfire MQ4) | 181.96 | — | Imported | 2026-05-06 |
| 27 | Qwen3.6-35B-A3B on NVIDIA GeForce RTX 5090 (ollama Q4_K_M) | 175.80 | — | Imported | 2026-05-06 |
| 28 | NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4 on NVIDIA H200 NVL (vLLM NVFP4) | 175.44 | — | Imported | 2026-05-06 |
| 29 | Qwen3.5-9B on NVIDIA RTX PRO 6000 Blackwell (vllm BF16) | 171.60 | — | Imported | 2026-05-06 |
| 30 | Qwen3.6-35B-A3B on 2x RTX 3090 (vllm INT4) | 170.50 | — | Imported | 2026-05-06 |
| 31 | Qwen3.6-27B-AWQ-BF16-INT4 on 2x RTX A6000 (vllm int4) | 165.56 | — | Imported | 2026-05-06 |
| 32 | Qwen3.6-35B-A3B-GGUF on NVIDIA GeForce RTX 4090 (llama.cpp Q4_K_M) | 165.06 | — | Imported | 2026-05-06 |
| 33 | Qwen3.6-35B-A3B-uncensored-heretic on RTX 4090 (llama.cpp Q3_K_M) | 164.09 | — | Imported | 2026-05-06 |
| 34 | gpt-oss-20b on 3x AMD Radeon AI Pro R9700 (llama.cpp Q8_0) | 160.82 | — | Imported | 2026-05-06 |
| 35 | gpt-oss-20b on 3x AMD Radeon AI Pro R9700 (llama.cpp Q8_0) | 159.72 | — | Imported | 2026-05-06 |
| 36 | Qwen3.6-35B-A3B on RTX 4090 (llama.cpp Q3_K_P) | 157.72 | — | Imported | 2026-05-06 |
| 37 | Qwen3.6-35B-A3B on RTX 4090 (llama.cpp Q3_K_P) | 157.72 | — | Imported | 2026-05-06 |
| 38 | Qwen3.5-35B-A3B-GGUF on RTX 4090 (llama.cpp UD-Q4_K_XL) | 157.40 | — | Imported | 2026-05-06 |
| 39 | Qwen3.6-27B on NVIDIA GeForce RTX 3090 Ti (llama.cpp UD-Q4_K_XL) | 155.94 | — | Imported | 2026-05-06 |
| 40 | Qwen3.5-35B-A3B on RX 7900 XTX (hipfire MQ4) | 154.62 | — | Imported | 2026-05-06 |
| 41 | Qwen3.6-35B-A3B on NVIDIA GeForce RTX 5080 (llama.cpp IQ3_XXS) | 150.60 | — | Imported | 2026-05-06 |
| 42 | Qwen3.6-27B-GGUF on NVIDIA GeForce RTX 3090 (llama.cpp UD-Q4_K_XL) | 150.07 | — | Imported | 2026-05-06 |
| 43 | Qwen3.5-0.8B-GGUF on 2x Tesla P100-PCIE-16GB (llama.cpp Q5_K_M) | 144.31 | — | Imported | 2026-05-06 |
| 44 | Nemotron-Cascade-2-30B-A3B on UNIFIED (mlx 4bit) | 140.80 | — | Imported | 2026-05-06 |
| 45 | Qwen3.5-35B-A3B on RX 7900 XTX (hipfire MQ4) | 140.59 | — | Imported | 2026-05-06 |
| 46 | Qwen3.6-27B on RTX 3090 Ti (llama.cpp Q4_K_M) | 140.28 | — | Imported | 2026-05-06 |
| 47 | Qwen3.6-35B-A3B on 2x NVIDIA GeForce RTX 3090 (vllm INT4) | 140.06 | — | Imported | 2026-05-06 |
| 48 | Qwen3.6-35B-A3B-FP8 on 2x RTX 3090 (vllm fp8) | 140.01 | — | Imported | 2026-05-06 |
| 49 | Ornstein3.6-35B-A3B-GGUF on 2x RTX 3090 (llama.cpp Q4_K_M) | 138.01 | — | Imported | 2026-05-06 |
| 50 | Qwen3.5-0.8B-GGUF on Tesla P100-PCIE-16GB (llama.cpp Q5_K_M) | 136.45 | — | Imported | 2026-05-06 |
| 51 | Qwen3.5-35B-A3B on RX 7900 XTX (hipfire MQ4) | 135.90 | — | Imported | 2026-05-06 |
| 52 | Qwen3.6-35B-A3B on RX 7900 XTX (hipfire MQ4) | 135.30 | — | Imported | 2026-05-06 |
| 53 | Ternary-Bonsai-8B-mlx-2bit on UNIFIED (mlx 2bit-ternary) | 135 | — | Imported | 2026-05-06 |
| 54 | Qwen3.6-35B-A3B on 2x NVIDIA GeForce RTX 3090 (vllm INT4) | 134.40 | — | Imported | 2026-05-06 |
| 55 | Ornstein3.6-35B-A3B on RX 7900 XTX (hipfire MQ4) | 134 | — | Imported | 2026-05-06 |
| 56 | Qwen3.6-27B-AWQ-BF16-INT4 on 2x NVIDIA RTX A6000 (vllm INT4) | 133.19 | — | Imported | 2026-05-06 |
| 57 | Qwen3-8B on RTX 3090 Ti (llama.cpp Q4_K_M) | 133.01 | — | Imported | 2026-05-06 |
| 58 | Qwen3.6-35B-A3B on 2x RTX 3090 Ti (llama.cpp UD-Q6_K_XL) | 131.58 | — | Imported | 2026-05-06 |
| 59 | Qwen3.6-27B-AWQ-BF16-INT4 on 2x RTX A6000 (vllm int4) | 129.19 | — | Imported | 2026-05-06 |
| 60 | Qwen3.6-35B-A3B-GGUF on 2x NVIDIA GeForce RTX 3090 (llama.cpp Q3_K_XL) | 125.94 | — | Imported | 2026-05-06 |
| 61 | Qwen3.5-35B-A3B-GGUF on RTX 3090 (llama.cpp UD-Q4_K_XL) | 124.40 | — | Imported | 2026-05-06 |
| 62 | Qwen3.5-9B-GGUF on RTX 5070 Ti (llama.cpp Q4_K_S) | 124.05 | — | Imported | 2026-05-06 |
| 63 | Qwen3.6-35B-A3B-GGUF on 2x RTX 3090 (llama.cpp UD-Q4_K_XL) | 124 | — | Imported | 2026-05-06 |
| 64 | Qwen3.5-9B on RX 7900 XTX (hipfire MQ4) | 122.30 | — | Imported | 2026-05-06 |
| 65 | Qwen3.5-35B-A3B on UNIFIED (ollama NVFP4) | 122.01 | — | Imported | 2026-05-06 |
| 66 | Qwen3.5-9B on RTX 3090 Ti (llama.cpp UD-Q4_K_XL) | 119.05 | — | Imported | 2026-05-06 |
| 67 | gemma-4-26B-A4B-it-MLX-4bit on UNIFIED (lmstudio 4bit) | 118.77 | — | Imported | 2026-05-06 |
| 68 | Qwen3.6-27B on RX 7900 XTX (hipfire MQ4) | 118.16 | — | Imported | 2026-05-06 |
| 69 | Qwen3.6-27B on RX 7900 XTX (hipfire MQ4) | 118.14 | — | Imported | 2026-05-06 |
| 70 | gemma-4-26b-a4b-it-4bit on UNIFIED (mlx 4bit) | 117 | — | Imported | 2026-05-06 |
| 71 | Carnice-9b-W8A16-AWQ on 2x RTX 3090 (vllm AWQ) | 114.26 | — | Imported | 2026-05-06 |
| 72 | Qwen3.6-35B-A3B on 3x AMD Radeon AI Pro R9700 (vllm MXFP4_MOE) | 112.76 | — | Imported | 2026-05-06 |
| 73 | Qwen3.5-9B-GGUF on 2x RTX 3090 (llama.cpp Q4_K_S) | 110.11 | — | Imported | 2026-05-06 |
| 74 | Qwen3.6-35B-A3B-4bit on UNIFIED (mlx 4bit) | 109 | — | Imported | 2026-05-06 |
| 75 | gemma-4-E2B-it on UNIFIED (llama.cpp Q4_K_M) | 107.07 | — | Imported | 2026-05-06 |
| 76 | gemma-4-E2B-it on UNIFIED (llama.cpp Q4_K_M) | 107 | — | Imported | 2026-05-06 |
| 77 | gemma-4-E2B-it on UNIFIED (llama.cpp Q4_K_M) | 106.60 | — | Imported | 2026-05-06 |
| 78 | gemma-4-E4B-it-GGUF on RTX 2080 Ti (llama.cpp Q4_K_M) | 106 | — | Imported | 2026-05-06 |
| 79 | Qwen3.6-35B-A3B on UNIFIED (lmstudio Q4_K_M) | 105.24 | — | Imported | 2026-05-06 |
| 80 | gemma-3-text-4b-it-4bit on UNIFIED (mlx 4bit) | 105.03 | — | Imported | 2026-05-06 |
| 81 | Qwen3.5-35B-A3B-4bit on UNIFIED (mlx 4bit) | 104.74 | — | Imported | 2026-05-06 |
| 82 | Qwen3.6-35B-A3B on AMD Radeon RX 7900 XTX (llama.cpp Q3_K_XL) | 104.53 | — | Imported | 2026-05-06 |
| 83 | Qwen3.6-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled on UNIFIED (lmstudio Q4_K_M) | 104.12 | — | Imported | 2026-05-06 |
| 84 | Qwen3.6-35B-A3B on UNIFIED (vllm NVFP4) | 102.05 | — | Imported | 2026-05-06 |
| 85 | Qwen3.6-35B-A3B-NVFP4 on UNIFIED (vllm NVFP4) | 102.05 | — | Imported | 2026-05-06 |
| 86 | Ornstein3.6-35B-A3B on NVIDIA GeForce RTX 3090 (llama.cpp Q4_K_M) | 102 | — | Imported | 2026-05-06 |
| 87 | Qwen3-30B-A3B on UNIFIED (llama.cpp Q4_K_M) | 101.95 | — | Imported | 2026-05-06 |
| 88 | Qwen3.6-27B-Text-NVFP4-MTP on RTX 5090 (vllm NVFP4) | 101.87 | — | Imported | 2026-05-06 |
| 89 | Qwen3-Coder-30B-A3B-Instruct-4bit on UNIFIED (mlx 4bit) | 101.05 | — | Imported | 2026-05-06 |
| 90 | Qwen3-Coder-30B-A3B-Instruct on UNIFIED (llama.cpp Q4_K_M) | 100 | — | Imported | 2026-05-06 |
| 91 | Qwen3-Coder-30B-A3B-Instruct on 3x AMD Radeon AI Pro R9700 (llama.cpp Q8_0) | 99.99 | — | Imported | 2026-05-06 |
| 92 | Qwen3.6-35B-A3B on 3x AMD Radeon AI Pro R9700 (vllm MXFP4_MOE) | 98.91 | — | Imported | 2026-05-06 |
| 93 | Huihui-Qwen3.6-27B-abliterated-NVFP4-MTP on RTX 5090 (vllm NVFP4) | 97.50 | — | Imported | 2026-05-06 |
| 94 | Qwen3.6-27B-AWQ-INT4 on 2x NVIDIA RTX A6000 (vllm INT4) | 97.30 | — | Imported | 2026-05-06 |
| 95 | Huihui-Qwen3.6-27B-abliterated-NVFP4-MTP on RTX 5090 (vllm NVFP4) | 97.10 | — | Imported | 2026-05-06 |
| 96 | gemma-4-26b-a4b-it-4bit on UNIFIED (mlx 4bit) | 96.71 | — | Imported | 2026-05-06 |
| 97 | Qwen3.6-35B-A3B on AMD Radeon RX 9070 XT (llama.cpp UD-Q2_K_XL) | 96.57 | — | Imported | 2026-05-06 |
| 98 | Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 on 3x AMD Radeon AI Pro R9700 (llama.cpp Q8_0) | 95.77 | — | Imported | 2026-05-06 |
| 99 | Nemotron-Cascade-2-30B-A3B on UNIFIED (llama.cpp Q4_K_M) | 95.41 | — | Imported | 2026-05-06 |
| 100 | Nemotron-Cascade-2-30B-A3B on 3x AMD Radeon AI Pro R9700 (llama.cpp Q8_0) | 95.39 | — | Imported | 2026-05-06 |
| 101 | Ministral-3-3B-Instruct-2512 on GTX 1080 Ti (llama.cpp Q4_K_M) | 94.90 | — | Imported | 2026-05-06 |
| 102 | Gemopus-4-26B-A4B-it-GGUF on AMD Radeon RX 7900 XTX (llama.cpp Q4_K_M) | 94.50 | — | Imported | 2026-05-06 |
| 103 | Huihui-Qwen3.6-27B-abliterated-NVFP4-MTP on RTX 5090 (vllm NVFP4) | 94.40 | — | Imported | 2026-05-06 |
| 104 | gemma-4-26B-A4B-it-GGUF on AMD Radeon RX 7900 XTX (llama.cpp Q4_K_S) | 94.30 | — | Imported | 2026-05-06 |
| 105 | GLM-4.7-Flash on 3x AMD Radeon AI Pro R9700 (llama.cpp Q8_0) | 93.28 | — | Imported | 2026-05-06 |
| 106 | Qwen3.5-9B-Claude-4.6-Opus-Reasoning-Distilled-GGUF on AMD Radeon RX 7900 XTX (llama.cpp Q4_K_S) | 93.20 | — | Imported | 2026-05-06 |
| 107 | GLM-4.7-Flash on 2x AMD Radeon AI Pro R9700 (llama.cpp MXFP4_MOE) | 92.93 | — | Imported | 2026-05-06 |
| 108 | MiniMax-M2.7-NVFP4 on 2x RTX PRO 6000 (sglang NVFP4) | 92.80 | — | Imported | 2026-05-06 |
| 109 | gemma-4-E4B-it-MLX-4bit on UNIFIED (lmstudio 4bit) | 91.40 | — | Imported | 2026-05-06 |
| 110 | Qwen3.6-27B-AEON-Ultimate-Uncensored-NVFP4 on UNIFIED (vllm NVFP4) | 90 | — | Imported | 2026-05-06 |
| 111 | Nemotron-Cascade-2-30B-A3B on 3x AMD Radeon AI Pro R9700 (llama.cpp Q8_0) | 89.78 | — | Imported | 2026-05-06 |
| 112 | Qwen3.6-35B-A3B on AMD Radeon RX 6800 (llama.cpp IQ3_S) | 89.15 | — | Imported | 2026-05-06 |
| 113 | gemma-4-E4B-it-GGUF on RTX 2080 Ti (llama.cpp Q6_K) | 89 | — | Imported | 2026-05-06 |
| 114 | Qwen3.6-27B-GGUF on RTX 5090 (llama.cpp Q4_K_M) | 89 | — | Imported | 2026-05-06 |
| 115 | Qwen3.6-27B-GGUF on RTX 5090 (llama.cpp Q4_K_M) | 89 | — | Imported | 2026-05-06 |
| 116 | Qwen3.6-27B-int4-AutoRound on NVIDIA GeForce RTX 3090 (vllm INT4) | 88.96 | — | Imported | 2026-05-06 |
| 117 | Qwen3-8B-GGUF on UNIFIED (llama.cpp Q4_K_M) | 88.25 | — | Imported | 2026-05-06 |
| 118 | Qwopus3.5-9B-v3 on AMD Radeon RX 9070 XT (llama.cpp Q4_K_M) | 87.49 | — | Imported | 2026-05-06 |
| 119 | Qwopus3.5-9B-v3 on AMD Radeon RX 9070 XT (llama.cpp Q4_K_M) | 87.49 | — | Imported | 2026-05-06 |
| 120 | Qwen3.5-9B on GTX 1080 Ti (llama.cpp Q4_K_M) | 87.40 | — | Imported | 2026-05-06 |
| 121 | gemma-4-26B-A4B-it on 2x AMD Radeon AI Pro R9700 (llama.cpp MXFP4_MOE) | 87.26 | — | Imported | 2026-05-06 |
| 122 | Qwen3-Coder-30B-A3B-Instruct on 3x AMD Radeon AI Pro R9700 (llama.cpp Q8_0) | 87.15 | — | Imported | 2026-05-06 |
| 123 | Qwen3.6-35B-A3B on AMD Radeon RX 7900 XTX (llama.cpp IQ4_XS) | 85.60 | — | Imported | 2026-05-06 |
| 124 | Qwen3.6-27B-int4-AutoRound on NVIDIA GeForce RTX 3090 (vllm INT4) | 85.22 | — | Imported | 2026-05-06 |
| 125 | Qwen3-Coder-Next-MLX-4bit on UNIFIED (lmstudio 4bit) | 85.16 | — | Imported | 2026-05-06 |
| 126 | Ornstein3.6-35B-A3B on 2x RTX 3090 (llama.cpp Q4_K_M) | 84.63 | — | Imported | 2026-05-06 |
| 127 | Qwen3.6-35B-A3B on UNIFIED (lmstudio 3bit-MLX) | 84.61 | — | Imported | 2026-05-06 |
| 128 | Qwen3.6-35B-A3B-UD-Q8_K_XL-mlx on UNIFIED (mlx Q8_K_XL) | 83.35 | — | Imported | 2026-05-06 |
| 129 | Huihui-GLM-4.7-Flash-abliterated on UNIFIED (llama.cpp Q8_0) | 82.57 | — | Imported | 2026-05-06 |
| 130 | Qwen3-Coder-Next on 2x AMD Radeon AI Pro R9700 (llama.cpp MXFP4_MOE) | 80.82 | — | Imported | 2026-05-06 |
| 131 | Qwen3-Coder-30B-A3B-Instruct on UNIFIED (llama.cpp Q5_K_M) | 80.46 | — | Imported | 2026-05-06 |
| 132 | Qwen3-Coder-30B-A3B-Instruct on UNIFIED (llama.cpp Q5_K_M) | 80.43 | — | Imported | 2026-05-06 |
| 133 | Qwen3-Coder-30B-A3B-Instruct on UNIFIED (llama.cpp Q5_K_M) | 80.41 | — | Imported | 2026-05-06 |
| 134 | Qwen3-Coder-Next on 3x AMD Radeon AI Pro R9700 (llama.cpp MXFP4_MOE) | 80.40 | — | Imported | 2026-05-06 |
| 135 | Qwen3-Coder-30B-A3B-Instruct on UNIFIED (llama.cpp Q5_K_M) | 80.11 | — | Imported | 2026-05-06 |
| 136 | gemma-4-E4B-it-GGUF on RTX 2080 Ti (llama.cpp Q8_0) | 80 | — | Imported | 2026-05-06 |
| 137 | Qwen3-Coder-30B-A3B-Instruct on UNIFIED (llama.cpp Q5_K_M) | 79.82 | — | Imported | 2026-05-06 |
| 138 | Qwen3-32B on 2x AMD Radeon AI Pro R9700 (vllm FP8) | 79.31 | — | Imported | 2026-05-06 |
| 139 | Qwen3.6-27B-int4-AutoRound on 2x RTX 3090 (vllm INT4) | 78.20 | — | Imported | 2026-05-06 |
| 140 | GLM-4.7-Flash-4bit on UNIFIED (mlx 4bit) | 77.53 | — | Imported | 2026-05-06 |
| 141 | Qwen2.5-7B-Instruct on RTX 4070 SUPER (llama.cpp Q4_K_M) | 77.40 | — | Imported | 2026-05-06 |
| 142 | Qwopus3.5-9B-v3 on AMD Radeon RX 9070 XT (llama.cpp Q4_K_M) | 77.09 | — | Imported | 2026-05-06 |
| 143 | Qwen3-Coder-Next-4bit on UNIFIED (mlx 4bit) | 76.45 | — | Imported | 2026-05-06 |
| 144 | Qwen3.6-35B-A3B-GGUF on AMD Radeon RX 9070 (llama.cpp IQ3_XS) | 76 | — | Imported | 2026-05-06 |
| 145 | Qwen3.6-35B-A3B on 3x AMD Radeon AI Pro R9700 (llama.cpp Q8_0) | 75.22 | — | Imported | 2026-05-06 |
| 146 | Qwen3.6-27B-FP8 on NVIDIA RTX PRO 6000 Blackwell (vllm FP8_E4M3) | 74.60 | — | Imported | 2026-05-06 |
| 147 | GLM-5.1-NVFP4-MTP on 8x NVIDIA RTX PRO 6000 Blackwell (sglang NVFP4) | 74.53 | — | Imported | 2026-05-06 |
| 148 | Kimi-K2.5 on 8x NVIDIA RTX PRO 6000 Blackwell (SGLang INT4) | 74 | — | Imported | 2026-05-06 |
| 149 | Qwopus-MoE-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 73.81 | — | Imported | 2026-05-06 |
| 150 | Qwopus-MoE-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 73.75 | — | Imported | 2026-05-06 |
| 151 | Carnice-MoE-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 72.71 | — | Imported | 2026-05-06 |
| 152 | Carnice-MoE-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 72.64 | — | Imported | 2026-05-06 |
| 153 | Carnice-MoE-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 72.59 | — | Imported | 2026-05-06 |
| 154 | Carnice-MoE-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 72.58 | — | Imported | 2026-05-06 |
| 155 | Qwopus-MoE-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 72.51 | — | Imported | 2026-05-06 |
| 156 | Qwopus-MoE-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 72.43 | — | Imported | 2026-05-06 |
| 157 | Qwopus-MoE-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 72.41 | — | Imported | 2026-05-06 |
| 158 | Qwopus-MoE-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 72.39 | — | Imported | 2026-05-06 |
| 159 | Qwen3.5-9B on 2x RTX 3090 (llama.cpp Q8_0) | 71.71 | — | Imported | 2026-05-06 |
| 160 | gpt-oss-120b on 3x AMD Radeon AI Pro R9700 (llama.cpp F16) | 71.53 | — | Imported | 2026-05-06 |
| 161 | Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive on UNIFIED (llama.cpp Q4_K_M) | 71.44 | — | Imported | 2026-05-06 |
| 162 | Qwen3.6-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 71.28 | — | Imported | 2026-05-06 |
| 163 | Ornstein3.6-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 71.22 | — | Imported | 2026-05-06 |
| 164 | Qwen3.6-27B-int4-AutoRound on 2x RTX 3090 (vllm INT4) | 71 | — | Imported | 2026-05-06 |
| 165 | Ornstein3.6-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 70.93 | — | Imported | 2026-05-06 |
| 166 | gpt-oss-120b on 3x AMD Radeon AI Pro R9700 (llama.cpp F16) | 70.72 | — | Imported | 2026-05-06 |
| 167 | Qwen3.6-35B-A3B on Intel Arc Pro B70 (llama.cpp Q4_K_M) | 70.35 | — | Imported | 2026-05-06 |
| 168 | Qwen3.6-27B-FP8 on 2x RTX 3090 Ti (vllm FP8) | 70.32 | — | Imported | 2026-05-06 |
| 169 | gpt-oss-120b on 3x AMD Radeon AI Pro R9700 (llama.cpp F16) | 69.99 | — | Imported | 2026-05-06 |
| 170 | Qwen3.6-35B-A3B on RX 7900 XTX (hipfire MQ4) | 68.60 | — | Imported | 2026-05-06 |
| 171 | Qwen3.6-27B-AWQ-BF16-INT4 on 4x RTX 3090 (vllm AWQ) | 67.09 | — | Imported | 2026-05-06 |
| 172 | Qwen3.6-35B-A3B on Tesla V100 SXM2 32GB (ollama Q4_K_M) | 66.82 | — | Imported | 2026-05-06 |
| 173 | supergemma4-26b-uncensored-gguf-v2 on UNIFIED (llama.cpp Q4_K_M) | 66.07 | — | Imported | 2026-05-06 |
| 174 | gemma-4-26B-A4B-it on UNIFIED (llama.cpp Q4_K_M) | 65.85 | — | Imported | 2026-05-06 |
| 175 | gemma-4-26B-A4B-it on UNIFIED (llama.cpp UD-Q4_K_M) | 65.71 | — | Imported | 2026-05-06 |
| 176 | Gemopus-4-26B-A4B-it on UNIFIED (llama.cpp Q4_K_M) | 64.27 | — | Imported | 2026-05-06 |
| 177 | Gemopus-4-26B-A4B-it on UNIFIED (llama.cpp Q4_K_M) | 64.07 | — | Imported | 2026-05-06 |
| 178 | Qwen3.5-4B on UNIFIED (llama.cpp IQ4_NL) | 63.92 | — | Imported | 2026-05-06 |
| 179 | gemma-4-26B-A4B-it on UNIFIED (llama.cpp UD-Q4_K_M) | 63.92 | — | Imported | 2026-05-06 |
| 180 | Qwen3.6-35B-A3B-GGUF on 2x Tesla P100-PCIE-16GB (llama.cpp Q5_K_S) | 63.91 | — | Imported | 2026-05-06 |
| 181 | Qwen3.6-35B-A3B on UNIFIED (llama.cpp UD-Q4_K_M) | 63.90 | — | Imported | 2026-05-06 |
| 182 | Qwen3.5-4B on UNIFIED (llama.cpp IQ4_NL) | 63.89 | — | Imported | 2026-05-06 |
| 183 | Qwen3.6-27B-int4-AutoRound on RTX 3090 (vllm INT4) | 63.80 | — | Imported | 2026-05-06 |
| 184 | Qwen3.5-4B on UNIFIED (llama.cpp Q4_K_M) | 62.41 | — | Imported | 2026-05-06 |
| 185 | Ornstein3.6-35B-A3B-GGUF on AMD Radeon RX 7900 XTX (llama.cpp Q4_K_S) | 62.40 | — | Imported | 2026-05-06 |
| 186 | Qwen3.6-27B-DFlash on UNIFIED (vllm FP8) | 62.13 | — | Imported | 2026-05-06 |
| 187 | Qwen3.6-35B-A3B on NVIDIA GeForce RTX 4070 (llama.cpp Q4_K_M) | 62.10 | — | Imported | 2026-05-06 |
| 188 | Qwen3.5-4B on UNIFIED (llama.cpp Q4_K_M) | 61.32 | — | Imported | 2026-05-06 |
| 189 | Qwen3.5-4B on UNIFIED (llama.cpp IQ4_NL) | 61.00 | — | Imported | 2026-05-06 |
| 190 | Qwen3.5-4B on UNIFIED (llama.cpp Q4_K_M) | 60.97 | — | Imported | 2026-05-06 |
| 191 | Qwen3.6-27B-AWQ-INT4 on 2x RTX 5060 Ti (vllm AWQ) | 60.90 | — | Imported | 2026-05-06 |
| 192 | gemma-4-26B-A4B-it on UNIFIED (llama.cpp Q4_K_M) | 60.74 | — | Imported | 2026-05-06 |
| 193 | gemma-4-26B-A4B-it on UNIFIED (llama.cpp Q4_K_M) | 60.57 | — | Imported | 2026-05-06 |
| 194 | gemma-4-26B-A4B-it on UNIFIED (llama.cpp Q4_K_M) | 60.57 | — | Imported | 2026-05-06 |
| 195 | Qwen3.5-35B-A3B-GGUF on UNIFIED (llama.cpp UD-Q4_K_XL) | 60.30 | — | Imported | 2026-05-06 |
| 196 | Qwen3.6-35B-A3B on NVIDIA GeForce RTX 4070 (llama.cpp Q4_K_M) | 60.20 | — | Imported | 2026-05-06 |
| 197 | Qwen3.6-35B-A3B on NVIDIA GeForce RTX 4060 Ti 16GB (llama.cpp IQ3_XXS) | 60.20 | — | Imported | 2026-05-06 |
| 198 | NVIDIA-Nemotron-3-Nano-Omni-30B-A3B-Reasoning on UNIFIED (llama.cpp UD-IQ4_NL) | 59.97 | — | Imported | 2026-05-06 |
| 199 | Qwen3.6-27B-AWQ-BF16-INT4-mtp-bf16 on 2x RTX 3090 (vllm AWQ_INT4) | 59.80 | — | Imported | 2026-05-06 |
| 200 | Ornstein3.6-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 59.08 | — | Imported | 2026-05-06 |
| 201 | Qwen3.6-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 58.75 | — | Imported | 2026-05-06 |
| 202 | Qwen3.6-35B-A3B on UNIFIED (llama.cpp Q4_K_M) | 58.73 | — | Imported | 2026-05-06 |
| 203 | NVIDIA-Nemotron-3-Nano-Omni-30B-A3B-Reasoning on UNIFIED (llama.cpp UD-IQ4_NL) | 58.51 | — | Imported | 2026-05-06 |
| 204 | Huihui-GLM-4.7-Flash-abliterated on UNIFIED (llama.cpp Q8_0) | 58.32 | — | Imported | 2026-05-06 |
| 205 | Qwen3.5-4B on UNIFIED (llama.cpp UD-Q4_K_XL) | 58.31 | — | Imported | 2026-05-06 |
| 206 | NVIDIA-Nemotron-3-Nano-Omni-30B-A3B-Reasoning on UNIFIED (llama.cpp UD-IQ4_NL) | 58.23 | — | Imported | 2026-05-06 |
| 207 | Qwen3.6-27B-int4-AutoRound on 2x RTX 3090 (vllm INT4) | 58.20 | — | Imported | 2026-05-06 |
| 208 | gemma-4-E4B-it on UNIFIED (llama.cpp Q4_K_M) | 57.73 | — | Imported | 2026-05-06 |
| 209 | Qwen3.6-35B-A3B on UNIFIED (llama.cpp UD-Q4_K_XL) | 57.62 | — | Imported | 2026-05-06 |
| 210 | Qwen2.5-7B-Instruct on GTX 1080 Ti (llama.cpp Q4_K_M) | 57.60 | — | Imported | 2026-05-06 |
| 211 | Qwen3.6-35B-A3B-GGUF on AMD Radeon RX 7900 XTX (llama.cpp Q4_K_S) | 57.60 | — | Imported | 2026-05-06 |
| 212 | Qwen3.6-35B-A3B on UNIFIED (llama.cpp UD-Q4_K_XL) | 57.59 | — | Imported | 2026-05-06 |
| 213 | Qwen3.6-27B-NVFP4 on RTX 5090 (vllm NVFP4) | 57.54 | — | Imported | 2026-05-06 |
| 214 | Qwen3.6-35B-A3B on UNIFIED (llama.cpp UD-Q4_K_XL) | 57.53 | — | Imported | 2026-05-06 |
| 215 | Qwen3.6-27B on NVIDIA GeForce RTX 5090 (llama.cpp Q4_K_XL) | 57.45 | — | Imported | 2026-05-06 |
| 216 | Qwen3.6-35B-A3B on UNIFIED (llama.cpp UD-Q4_K_XL) | 57.36 | — | Imported | 2026-05-06 |
| 217 | gemma-4-E4B-it on UNIFIED (llama.cpp Q4_K_M) | 57.25 | — | Imported | 2026-05-06 |
| 218 | Qwen3-Coder-Next on UNIFIED (llama.cpp Q5_K_XL) | 57.14 | — | Imported | 2026-05-06 |
| 219 | Qwen3.5-4B on UNIFIED (llama.cpp IQ4_NL) | 57.12 | — | Imported | 2026-05-06 |
| 220 | Qwen3.6-35B-A3B-GGUF-Strix on UNIFIED (llama.cpp DYNAMIC) | 56.77 | — | Imported | 2026-05-06 |
| 221 | gemma-4-E4B-it on UNIFIED (llama.cpp Q4_K_M) | 56.74 | — | Imported | 2026-05-06 |
| 222 | Qwen3.6-35B-A3B on UNIFIED (llama.cpp UD-Q4_K_XL) | 56.67 | — | Imported | 2026-05-06 |
| 223 | Qwen3-VL-30B-A3B-Instruct on UNIFIED (llama.cpp Q8_0) | 56.62 | — | Imported | 2026-05-06 |
| 224 | Qwen3.5-35B-A3B on UNIFIED (llama.cpp UD-Q4_K_XL) | 56.06 | — | Imported | 2026-05-06 |
| 225 | Qwen3-Coder-Next on UNIFIED (llama.cpp Q4_K_M) | 55.80 | — | Imported | 2026-05-06 |
| 226 | Qwen3-Coder-Next on UNIFIED (llama.cpp Q4_K_M) | 55.77 | — | Imported | 2026-05-06 |
| 227 | Qwen3-Coder-Next on UNIFIED (llama.cpp Q4_K_M) | 55.77 | — | Imported | 2026-05-06 |
| 228 | Qwen3.6-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled-GGUF on UNIFIED (llama.cpp Q6_K) | 55.65 | — | Imported | 2026-05-06 |
| 229 | Qwen3-Coder-Next on UNIFIED (llama.cpp Q4_K_M) | 55.60 | — | Imported | 2026-05-06 |
| 230 | gemma-4-26B-A4B-it on UNIFIED (llama.cpp Q6_K) | 55.52 | — | Imported | 2026-05-06 |
| 231 | Qwen3.6-27B-AWQ-BF16-INT4-mtp-bf16 on 2x RTX 3090 (vllm AWQ) | 55.20 | — | Imported | 2026-05-06 |
| 232 | gemma-4-26B-A4B-it on 2x Tesla P100-PCIE-16GB (llama.cpp Q5_K_M) | 54.78 | — | Imported | 2026-05-06 |
| 233 | gemma-4-31B-it on UNIFIED (llama.cpp Q4_K_M) | 54.73 | — | Imported | 2026-05-06 |
| 234 | Qwen3.6-35B-A3B on NVIDIA GeForce RTX 4070 (llama.cpp Q4_K_M) | 54.50 | — | Imported | 2026-05-06 |
| 235 | Qwen3.6-35B-A3B on NVIDIA GeForce RTX 4070 (llama.cpp Q4_K_M) | 54.50 | — | Imported | 2026-05-06 |
| 236 | Qwopus-MoE-35B-A3B on UNIFIED (llama.cpp Q8_0) | 53.16 | — | Imported | 2026-05-06 |
| 237 | Qwopus-MoE-35B-A3B on UNIFIED (llama.cpp Q8_0) | 53.14 | — | Imported | 2026-05-06 |
| 238 | Hermes-3-Llama-3.1-8B on GTX 1080 Ti (llama.cpp Q4_K_M) | 53.10 | — | Imported | 2026-05-06 |
| 239 | Qwopus-MoE-35B-A3B on UNIFIED (llama.cpp Q8_0) | 53.08 | — | Imported | 2026-05-06 |
| 240 | Qwen3.6-35B-A3B on UNIFIED (llama.cpp Q8_0) | 53.05 | — | Imported | 2026-05-06 |
| 241 | Qwen3.6-35B-A3B on UNIFIED (llama.cpp Q8_0) | 52.81 | — | Imported | 2026-05-06 |
| 242 | Llama-3.1-8B-Instruct on GTX 1080 Ti (llama.cpp Q4_K_M) | 52.80 | — | Imported | 2026-05-06 |
| 243 | Qwen3.6-35B-A3B-GGUF on UNIFIED (llama.cpp UD-Q4_K_XL) | 52.12 | — | Imported | 2026-05-06 |
| 244 | Qwen3.5-35B-A3B on UNIFIED (llama.cpp Q8_0) | 51.68 | — | Imported | 2026-05-06 |
| 245 | Qwen3.5-35B-A3B-FP8 on UNIFIED (vllm fp8) | 51.40 | — | Imported | 2026-05-06 |
| 246 | Qwen3-Coder-Next on UNIFIED (llama.cpp Q5_K_M) | 51.39 | — | Imported | 2026-05-06 |
| 247 | gemma-4-26B-A4B-it on UNIFIED (llama.cpp UD-Q4_K_M) | 51.22 | — | Imported | 2026-05-06 |
| 248 | Qwen3-Coder-Next on UNIFIED (llama.cpp Q5_K_M) | 51.21 | — | Imported | 2026-05-06 |
| 249 | Qwen3-Coder-Next on UNIFIED (llama.cpp Q5_K_M) | 51.18 | — | Imported | 2026-05-06 |
| 250 | Qwen3-Coder-Next on UNIFIED (llama.cpp Q5_K_M) | 51.18 | — | Imported | 2026-05-06 |
| 251 | Qwen3-8B on GTX 1080 Ti (llama.cpp Q4_K_M) | 50.70 | — | Imported | 2026-05-06 |
| 252 | gemma-4-26B-A4B-it-GGUF on UNIFIED (llama.cpp UD-Q4_K_XL) | 50.70 | — | Imported | 2026-05-06 |
| 253 | Qwen3.6-35B-A3B on UNIFIED (llama.cpp UD-Q4_K_M) | 50.48 | — | Imported | 2026-05-06 |
| 254 | Qwen3.6-35B-A3B-GGUF on UNIFIED (llama.cpp UD-Q6_K_XL) | 50.36 | — | Imported | 2026-05-06 |
| 255 | Qwopus-MoE-35B-A3B on UNIFIED (llama.cpp Q8_0) | 50.27 | — | Imported | 2026-05-06 |
| 256 | Qwen3.5-9B-GGUF on 2x Tesla P100-PCIE-16GB (llama.cpp Q5_K_M) | 49.90 | — | Imported | 2026-05-06 |
| 257 | Qwen3.6-27B on RTX 4090 (llama.cpp Q3_K_XL) | 49.37 | — | Imported | 2026-05-06 |
| 258 | Qwen3.6-27B-int4-AutoRound on 2x Intel Arc Pro B70 32GB (vllm INT4 AutoRound W4A16) | 49.10 | — | Imported | 2026-05-06 |
| 259 | gemma-4-26B-A4B-it on UNIFIED (llama.cpp UD-Q4_K_M) | 48.65 | — | Imported | 2026-05-06 |
| 260 | Qwen3.6-27B-FP8 on 2x RTX 3090 (vllm fp8) | 48.50 | — | Imported | 2026-05-06 |
| 261 | gemma-4-26B-A4B-it on UNIFIED (llama.cpp UD-Q4_K_M) | 48.33 | — | Imported | 2026-05-06 |
| 262 | Qwen3.6-27B-int4-AutoRound on 2x Intel Arc Pro B70 32GB (vllm INT4 AutoRound W4A16) | 48.30 | — | Imported | 2026-05-06 |
| 263 | gpt-oss-20b on 2x AMD Radeon AI Pro R9700 (vllm MXFP4_MOE) | 48.28 | — | Imported | 2026-05-06 |
| 264 | Qwen3-30B-A3B-Instruct-2507-AWQ-4bit-Q4_K_M-GGUF on AMD Radeon RX 7900 XTX (llama.cpp Q4_K_M) | 47.80 | — | Imported | 2026-05-06 |
| 265 | Qwen3-VL-30B-A3B-Instruct on UNIFIED (llama.cpp Q8_0) | 47.72 | — | Imported | 2026-05-06 |
| 266 | Qwen3.6-27B-FP8 on 4x Intel Arc Pro B70 (vllm fp8) | 47.67 | — | Imported | 2026-05-06 |
| 267 | gemma-4-26B-A4B-it on UNIFIED (llama.cpp Q8_0) | 47.16 | — | Imported | 2026-05-06 |
| 268 | Qwen3-Coder-30B-A3B-Instruct-GGUF on AMD Radeon RX 7900 XTX (llama.cpp Q4_K_M) | 46.90 | — | Imported | 2026-05-06 |
| 269 | Qwen3.5-35B-A3B on UNIFIED (llama.cpp UD-Q4_K_XL) | 46.90 | — | Imported | 2026-05-06 |
| 270 | gemma-4-31B-it on NVIDIA RTX PRO 6000 Blackwell Workstation Edition (vllm BF16) | 46.80 | — | Imported | 2026-05-06 |
| 271 | Qwen3.5-27B-GGUF on RTX 4090 (llama.cpp Q4_K_M) | 46.69 | — | Imported | 2026-05-06 |
| 272 | Qwen3.6-27B on RTX 4090 (llama.cpp Q4_K_M) | 46.64 | — | Imported | 2026-05-06 |
| 273 | Qwen3.6-35B-A3B on UNIFIED (llama.cpp Q8_0) | 46.25 | — | Imported | 2026-05-06 |
| 274 | Qwen3.6-27B on 3x Intel Arc Pro B70 (llama.cpp Q4_0) | 46.12 | — | Imported | 2026-05-06 |
| 275 | Qwen3.6-27B-FP8 on 4x Intel Arc Pro B70 32GB (vllm compressed-tensors FP8) | 46.07 | — | Imported | 2026-05-06 |
| 276 | Gemopus-4-26B-A4B-it on UNIFIED (llama.cpp Q8_0) | 45.96 | — | Imported | 2026-05-06 |
| 277 | Gemopus-4-26B-A4B-it on UNIFIED (llama.cpp Q8_0) | 45.70 | — | Imported | 2026-05-06 |
| 278 | Qwen3.6-27B on 3x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 45.62 | — | Imported | 2026-05-06 |
| 279 | Qwen3.5-27B-int4-AutoRound on 2x RTX 3090 (vllm INT4) | 45.55 | — | Imported | 2026-05-06 |
| 280 | Qwen3.6-27B-int4-AutoRound on Intel Arc Pro B70 32GB (vllm INT4 AutoRound W4A16) | 45.20 | — | Imported | 2026-05-06 |
| 281 | Qwen3.5-35B-A3B on UNIFIED (llama.cpp Q8_0) | 45.08 | — | Imported | 2026-05-06 |
| 282 | gemma-4-26B-A4B-it on UNIFIED (llama.cpp UD-Q4_K_M) | 44.91 | — | Imported | 2026-05-06 |
| 283 | Qwen3.6-27B on 3x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 44.81 | — | Imported | 2026-05-06 |
| 284 | Qwen3.6-35B-A3B on UNIFIED (llama.cpp UD-Q4_K_M) | 44.73 | — | Imported | 2026-05-06 |
| 285 | Qwen3.6-27B on 3x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 44.24 | — | Imported | 2026-05-06 |
| 286 | Qwen3.6-27B on RTX 4090 (llama.cpp IQ4_NL) | 44.20 | — | Imported | 2026-05-06 |
| 287 | Qwen3.6-27B on 3x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 44.18 | — | Imported | 2026-05-06 |
| 288 | Qwen3.6-27B-FP8 on 2x RTX 3090 (vllm fp8) | 44.17 | — | Imported | 2026-05-06 |
| 289 | Qwen3.6-27B on 3x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 44.00 | — | Imported | 2026-05-06 |
| 290 | Qwen3.5-27B-GGUF on RTX 4090 (llama.cpp Q4_0) | 44 | — | Imported | 2026-05-06 |
| 291 | Qwen3.5-27B on RX 7900 XTX (hipfire MQ4) | 43.70 | — | Imported | 2026-05-06 |
| 292 | Qwen3.6-27B-FP8 on 2x RTX 3090 (vllm fp8) | 43.62 | — | Imported | 2026-05-06 |
| 293 | Qwen3.6-27B on 3x Intel Arc Pro B70 (llama.cpp Q4_0) | 43.61 | — | Imported | 2026-05-06 |
| 294 | Qwen3.6-27B on RX 7900 XTX (hipfire MQ4) | 43.60 | — | Imported | 2026-05-06 |
| 295 | Qwen3.6-27B-AWQ-BF16-INT4 on 2x RTX 3090 (vllm AWQ) | 43.19 | — | Imported | 2026-05-06 |
| 296 | Qwen3.6-27B-AWQ-BF16-INT4 on 2x RTX 3090 (vllm AWQ) | 43.17 | — | Imported | 2026-05-06 |
| 297 | Qwen3.6-27B on RTX 4090 (llama.cpp Q4_K_P) | 42.70 | — | Imported | 2026-05-06 |
| 298 | Qwen3.6-27B-FP8 on 4x Intel Arc Pro B70 32GB (vllm compressed-tensors FP8) | 42.49 | — | Imported | 2026-05-06 |
| 299 | Qwen3.6-27B on 3x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 42.43 | — | Imported | 2026-05-06 |
| 300 | Qwen3.5-9B on GTX 1080 Ti (llama.cpp Q4_K_M) | 42.30 | — | Imported | 2026-05-06 |
| 301 | Qwen3.6-27B on 2x NVIDIA GeForce RTX 3090 (llama.cpp IQ4_NL) | 42.07 | — | Imported | 2026-05-06 |
| 302 | Qwen3.6-27B-GGUF on RTX 4090 (llama.cpp UD-Q4_K_XL) | 41.82 | — | Imported | 2026-05-06 |
| 303 | Qwen3.6-27B-DFlash on 3x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 41.74 | — | Imported | 2026-05-06 |
| 304 | Qwen3.6-27B on 3x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 41.66 | — | Imported | 2026-05-06 |
| 305 | Qwen3.6-27B-FP8 on 4x Intel Arc Pro B70 32GB (vllm FP8) | 41.50 | — | Imported | 2026-05-06 |
| 306 | Qwen3.6-35B-A3B-GGUF on UNIFIED (llama.cpp UD-Q8_K_XL) | 41.40 | — | Imported | 2026-05-06 |
| 307 | Qwen3.6-27B-DFlash on 3x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 41.37 | — | Imported | 2026-05-06 |
| 308 | Qwen3.6-27B-int4-AutoRound on Intel Arc Pro B70 32GB (vllm INT4 AutoRound W4A16) | 41.30 | — | Imported | 2026-05-06 |
| 309 | Qwen2.5-7B-Instruct on GTX 1080 Ti (llama.cpp Q5_K_M) | 41.10 | — | Imported | 2026-05-06 |
| 310 | Qwen3.6-27B on 2x RTX 3090 (vllm fp8) | 40.86 | — | Imported | 2026-05-06 |
| 311 | Qwen3.5-27B-GGUF on RTX 3090 (llama.cpp Q4_K_M) | 40.85 | — | Imported | 2026-05-06 |
| 312 | Qwen3.6-27B on 2x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 40.49 | — | Imported | 2026-05-06 |
| 313 | Qwen3.6-27B on 2x NVIDIA GeForce RTX 3090 (llama.cpp IQ4_NL) | 40.39 | — | Imported | 2026-05-06 |
| 314 | Qwen3.6-35B-A3B-GGUF on 2x RTX 3090 (llama.cpp UD-Q6_K_XL) | 40.34 | — | Imported | 2026-05-06 |
| 315 | Qwen3.6-27B on 2x NVIDIA GeForce RTX 3090 (llama.cpp IQ4_NL) | 40.17 | — | Imported | 2026-05-06 |
| 316 | Qwen3.6-27B-DFlash on 2x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 39.85 | — | Imported | 2026-05-06 |
| 317 | Huihui-GLM-4.7-Flash-abliterated on UNIFIED (llama.cpp Q8_0) | 39.69 | — | Imported | 2026-05-06 |
| 318 | Qwen3.6-27B-FP8 on 4x Intel Arc Pro B70 32GB (vllm FP8) | 39.26 | — | Imported | 2026-05-06 |
| 319 | gemma-4-31B-it-GGUF on RTX 3090 (llama.cpp Q4_K_M) | 38.90 | — | Imported | 2026-05-06 |
| 320 | gemma-4-26B-A4B-it-GGUF on UNIFIED (llama.cpp UD-Q8_K_XL) | 38.80 | — | Imported | 2026-05-06 |
| 321 | Qwen3.6-27B on NVIDIA GeForce RTX 3090 (llama.cpp UD-Q4_K_XL) | 38.78 | — | Imported | 2026-05-06 |
| 322 | gemma-4-31B-it-GGUF on RTX 3090 (llama.cpp Q4_K_M) | 38.73 | — | Imported | 2026-05-06 |
| 323 | Hermes-3-Llama-3.1-8B on UNIFIED (llama.cpp Q5_K_M) | 38.63 | — | Imported | 2026-05-06 |
| 324 | Qwen3.6-27B-DFlash on 2x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 38.62 | — | Imported | 2026-05-06 |
| 325 | Hermes-3-Llama-3.1-8B on UNIFIED (llama.cpp Q5_K_M) | 38.53 | — | Imported | 2026-05-06 |
| 326 | Qwen3.5-9B on UNIFIED (llama.cpp Q4_K_M) | 38.52 | — | Imported | 2026-05-06 |
| 327 | Qwen3.6-27B-GGUF on 2x RTX 3090 (llama.cpp Q4_K_P) | 38.41 | — | Imported | 2026-05-06 |
| 328 | Qwen3.5-27B-GGUF on RTX 3090 (llama.cpp Q4_0) | 38.40 | — | Imported | 2026-05-06 |
| 329 | Qwen3.6-27B-DFlash on 3x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 38.36 | — | Imported | 2026-05-06 |
| 330 | Hermes-3-Llama-3.1-8B on UNIFIED (llama.cpp Q5_K_M) | 38.29 | — | Imported | 2026-05-06 |
| 331 | Hermes-3-Llama-3.1-8B on GTX 1080 Ti (llama.cpp Q5_K_M) | 38.20 | — | Imported | 2026-05-06 |
| 332 | Llama-3.1-8B-Instruct on GTX 1080 Ti (llama.cpp Q5_K_M) | 38.10 | — | Imported | 2026-05-06 |
| 333 | Carnice-9b on UNIFIED (llama.cpp Q4_K_M) | 38.07 | — | Imported | 2026-05-06 |
| 334 | Qwen3.5-9B on UNIFIED (llama.cpp IQ4_NL) | 38.01 | — | Imported | 2026-05-06 |
| 335 | Carnice-9b on UNIFIED (llama.cpp Q4_K_M) | 37.93 | — | Imported | 2026-05-06 |
| 336 | Qwen3.6-27B on NVIDIA GeForce RTX 3090 (llama.cpp UD-Q4_K_XL) | 37.88 | — | Imported | 2026-05-06 |
| 337 | Qwen3.6-27B-GGUF on 2x NVIDIA GeForce RTX 3090 (llama.cpp Q4_K_M) | 37.78 | — | Imported | 2026-05-06 |
| 338 | Qwen3.6-27B-DFlash on 2x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 37.69 | — | Imported | 2026-05-06 |
| 339 | Qwen3.5-9B on UNIFIED (llama.cpp IQ4_NL) | 37.68 | — | Imported | 2026-05-06 |
| 340 | Qwen3.5-9B on UNIFIED (llama.cpp IQ4_NL) | 37.58 | — | Imported | 2026-05-06 |
| 341 | Qwen3.5-9B on UNIFIED (llama.cpp IQ4_NL) | 37.44 | — | Imported | 2026-05-06 |
| 342 | Qwen3.5-9B on UNIFIED (llama.cpp Q4_K_M) | 37.24 | — | Imported | 2026-05-06 |
| 343 | Qwen3.5-9B on UNIFIED (llama.cpp Q4_K_M) | 37.21 | — | Imported | 2026-05-06 |
| 344 | Qwen3.6-27B-AEON-Ultimate-Uncensored-NVFP4 on UNIFIED (vllm nvfp4) | 37 | — | Imported | 2026-05-06 |
| 345 | Gemma-4-26B-A4B-NVFP4 on NVIDIA RTX PRO 6000 Blackwell Workstation Edition (vllm NVFP4) | 36.90 | — | Imported | 2026-05-06 |
| 346 | Qwen3-8B on GTX 1080 Ti (llama.cpp Q5_K_M) | 36.80 | — | Imported | 2026-05-06 |
| 347 | Qwen3.6-27B-GGUF on NVIDIA GeForce RTX 3090 Ti (llama.cpp Q4_K_M) | 36.37 | — | Imported | 2026-05-06 |
| 348 | Llama-3.1-8B-Instruct on GTX 1080 Ti (llama.cpp unknown) | 36.30 | — | Imported | 2026-05-06 |
| 349 | Qwen3.6-27B-GGUF on NVIDIA GeForce RTX 3090 Ti (llama.cpp Q4_K_M) | 35.71 | — | Imported | 2026-05-06 |
| 350 | Qwen3.6-27B-int4-AutoRound on 2x Intel Arc Pro B70 32GB (vllm INT4 AutoRound W4A16) | 35.60 | — | Imported | 2026-05-06 |
| 351 | Carnice-9b on UNIFIED (llama.cpp Q4_K_M) | 35.40 | — | Imported | 2026-05-06 |
| 352 | Carnice-9b on UNIFIED (llama.cpp Q4_K_M) | 35.36 | — | Imported | 2026-05-06 |
| 353 | Qwen3.5-35B-A3B-GGUF on UNIFIED (llama.cpp UD-Q8_K_XL) | 35.10 | — | Imported | 2026-05-06 |
| 354 | Qwen3.5-9B on UNIFIED (llama.cpp UD-Q4_K_XL) | 35.00 | — | Imported | 2026-05-06 |
| 355 | Qwen3.6-35B-A3B on NVIDIA GeForce RTX 3060 (llama.cpp Q4_K_M) | 35 | — | Imported | 2026-05-06 |
| 356 | Qwen3.6-27B on 4x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 34.38 | — | Imported | 2026-05-06 |
| 357 | Qwen3.5-9B on GTX 1080 Ti (llama.cpp Q5_K_M) | 34.30 | — | Imported | 2026-05-06 |
| 358 | Qwen3.6-35B-A3B on AMD Radeon RX 9070 (llama.cpp Q4_K_XL) | 34.20 | — | Imported | 2026-05-06 |
| 359 | Qwen3.5-27B-AWQ-BF16-INT4 on 2x Multi-GPU (vllm AWQ) | 33.80 | — | Imported | 2026-05-06 |
| 360 | Qwen3.6-35B-A3B on 2x AMD Radeon AI Pro R9700 (vllm MXFP4_MOE) | 33.56 | — | Imported | 2026-05-06 |
| 361 | Qwen3.6-35B-A3B on NVIDIA GeForce RTX 3070 Ti (llama.cpp IQ3_S) | 33.45 | — | Imported | 2026-05-06 |
| 362 | Qwen3.6-27B-GGUF on AMD Radeon RX 7900 XTX (llama.cpp Q3_K_XL) | 33.31 | — | Imported | 2026-05-06 |
| 363 | Qwen3.6-27B-FP8 on 4x Intel Arc Pro B70 32GB (vllm FP8 W8A8 compressed-tensors) | 33.08 | — | Imported | 2026-05-06 |
| 364 | Qwen3.5-9B-4bit on UNIFIED (mlx int4) | 33 | — | Imported | 2026-05-06 |
| 365 | Qwen3.6-27B-MXFP4-CRACK on UNIFIED (mlx MXFP4) | 33 | — | Imported | 2026-05-06 |
| 366 | Qwen3.6-27B on UNIFIED (vllm NVFP4) | 32.83 | — | Imported | 2026-05-06 |
| 367 | Qwen3.6-27B-NVFP4 on UNIFIED (vllm NVFP4) | 32.83 | — | Imported | 2026-05-06 |
| 368 | Qwen3.6-27B-GGUF on NVIDIA GeForce RTX 3090 Ti (llama.cpp Q4_K_M) | 32.65 | — | Imported | 2026-05-06 |
| 369 | Qwen3.5-9B-GGUF on Tesla P100-PCIE-16GB (llama.cpp Q5_K_M) | 32.35 | — | Imported | 2026-05-06 |
| 370 | Qwen3.6-27B-NVFP4 on UNIFIED (vllm NVFP4) | 32.17 | — | Imported | 2026-05-06 |
| 371 | Qwen3.6-27B-AWQ-INT4 on 2x RTX 5060 Ti (vllm AWQ) | 32.10 | — | Imported | 2026-05-06 |
| 372 | Qwen3.6-27B on 4x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 31.91 | — | Imported | 2026-05-06 |
| 373 | Qwen3.6-27B-int4-AutoRound on Intel Arc Pro B70 32GB (vllm INT4 AutoRound W4A16) | 31.80 | — | Imported | 2026-05-06 |
| 374 | Qwen3.6-35B-A3B on 2x AMD Radeon AI Pro R9700 (vllm MXFP4_MOE) | 31.70 | — | Imported | 2026-05-06 |
| 375 | Qwen3.6-27B-DFlash on 4x Intel Arc Pro B70 32GB (llama.cpp Q4_0) | 31.48 | — | Imported | 2026-05-06 |
| 376 | MiniMax-M2.7-JANGTQ-CRACK on UNIFIED (mlx JANGTQ) | 31 | — | Imported | 2026-05-06 |
| 377 | MiniMax-M2.7 on UNIFIED (llama.cpp UD-IQ4_XS) | 30.98 | — | Imported | 2026-05-06 |
| 378 | MiniMax-M2.7 on UNIFIED (llama.cpp UD-IQ4_XS) | 30.88 | — | Imported | 2026-05-06 |
| 379 | MiniMax-M2.7 on UNIFIED (llama.cpp UD-IQ4_XS) | 30.88 | — | Imported | 2026-05-06 |
| 380 | Qwen3.5-4B on GTX 1650 (llama.cpp IQ4_NL) | 30.58 | — | Imported | 2026-05-06 |
| 381 | Qwen3.6-27B-AEON-Ultimate-Uncensored-NVFP4 on UNIFIED (vllm NVFP4) | 30 | — | Imported | 2026-05-06 |
| 382 | MiniMax-M2.7 on UNIFIED (llama.cpp UD-Q3_K_S) | 28.94 | — | Imported | 2026-05-06 |
| 383 | Qwen3.6-27B-GGUF on Intel Arc Pro B70 (llama.cpp Q4_0) | 28.80 | — | Imported | 2026-05-06 |
| 384 | Qwopus3.5-27B-v3-GGUF on AMD Radeon RX 7900 XTX (llama.cpp Q4_K_S) | 28.60 | — | Imported | 2026-05-06 |
| 385 | Qwen3.6-27B-GGUF on AMD Radeon RX 7900 XTX (llama.cpp Q4_K_S) | 28.50 | — | Imported | 2026-05-06 |
| 386 | Qwen3.5-122B-A10B-int4-AutoRound on UNIFIED (vllm int4) | 28.10 | — | Imported | 2026-05-06 |
| 387 | Qwen3.6-27B-FP8 on 4x Intel Arc Pro B70 32GB (vllm FP8 compressed-tensors) | 28.04 | — | Imported | 2026-05-06 |
| 388 | Qwen3-VL-2B-Instruct on UNIFIED (llama.cpp Q4_K_M) | 27.92 | — | Imported | 2026-05-06 |
| 389 | Qwen3.6-27B-GGUF on Intel Arc Pro B70 (llama.cpp Q4_0) | 27.79 | — | Imported | 2026-05-06 |
| 390 | Qwen3.5-27B-GGUF on 2x Multi-GPU (llama.cpp Q8_0) | 27.74 | — | Imported | 2026-05-06 |
| 391 | Qwen3.6-27B-GGUF on Intel Arc Pro B70 (llama.cpp Q4_0) | 27.65 | — | Imported | 2026-05-06 |
| 392 | Qwen3.6-35B-A3B on GTX 1080 Ti (llama.cpp IQ2_XXS) | 27.60 | — | Imported | 2026-05-06 |
| 393 | Qwen3.6-27B on UNIFIED (llama.cpp Q4_K_M) | 27.45 | — | Imported | 2026-05-06 |
| 394 | Qwen3.5-122B-A10B on 3x AMD Radeon AI Pro R9700 (llama.cpp MXFP4_MOE) | 27.31 | — | Imported | 2026-05-06 |
| 395 | Qwen3.5-122B-A10B on 3x AMD Radeon AI Pro R9700 (llama.cpp MXFP4_MOE) | 27.19 | — | Imported | 2026-05-06 |
| 396 | Qwen3.6-27B-GGUF on Intel Arc Pro B70 (llama.cpp Q4_0) | 27.03 | — | Imported | 2026-05-06 |
| 397 | Qwen3.6-35B-A3B on UNIFIED (vllm AWQ) | 26.89 | — | Imported | 2026-05-06 |
| 398 | Qwen3.6-27B-DFlash on 2x Intel Arc Pro B70 (llama.cpp Q4_0) | 26.87 | — | Imported | 2026-05-06 |
| 399 | Qwen3.5-122B-A10B on 3x AMD Radeon AI Pro R9700 (llama.cpp MXFP4_MOE) | 26.84 | — | Imported | 2026-05-06 |
| 400 | Qwen3.6-27B-GGUF on 2x RTX 3090 (llama.cpp Q8_0) | 26.82 | — | Imported | 2026-05-06 |
| 401 | Qwen3.5-122B-A10B on 3x AMD Radeon AI Pro R9700 (llama.cpp MXFP4_MOE) | 26.81 | — | Imported | 2026-05-06 |
| 402 | Qwen3.6-27B-GGUF on Intel Arc Pro B70 (llama.cpp Q4_K_S) | 26.68 | — | Imported | 2026-05-06 |
| 403 | Qwen3.6-27B on UNIFIED (llama.cpp Q4_K_M) | 26.34 | — | Imported | 2026-05-06 |
| 404 | Qwen3.6-27B on UNIFIED (llama.cpp Q4_K_M) | 26.34 | — | Imported | 2026-05-06 |
| 405 | Qwen3.5-122B-A10B-REAP-20-GGUF on UNIFIED (llama.cpp Q4_K_M-REAP-20) | 26.25 | — | Imported | 2026-05-06 |
| 406 | Qwen3.6-27B-GGUF on Intel Arc Pro B70 (llama.cpp IQ4_NL) | 26.04 | — | Imported | 2026-05-06 |
| 407 | Qwen3.6-35B-A3B on GTX 1080 Ti (llama.cpp IQ2_XS) | 26 | — | Imported | 2026-05-06 |
| 408 | Qwen3.6-27B on 2x Intel Arc Pro B70 32GB (llama.cpp Q8_0) | 25.73 | — | Imported | 2026-05-06 |
| 409 | Qwen3.6-27B-GGUF on Intel Arc Pro B70 (llama.cpp Q4_0) | 25.16 | — | Imported | 2026-05-06 |
| 410 | Qwen3.6-27B-GGUF on Intel Arc Pro B70 (llama.cpp Q4_K_M) | 24.93 | — | Imported | 2026-05-06 |
| 411 | Qwen3.6-27B on AMD Radeon RX 9070 XT (llama.cpp UD-IQ3_XXS) | 24.91 | — | Imported | 2026-05-06 |
| 412 | nvidia_Nemotron-Cascade-2-30B-A3B-GGUF on UNIFIED (llama.cpp Q4_K_M) | 24.75 | — | Imported | 2026-05-06 |
| 413 | Qwen3.6-35B-A3B-GGUF on Tesla P100-PCIE-16GB (llama.cpp IQ3_S) | 24.17 | — | Imported | 2026-05-06 |
| 414 | Qwen3.6-27B-GGUF on 2x RTX 3090 (llama.cpp UD-Q4_K_XL) | 24.16 | — | Imported | 2026-05-06 |
| 415 | Qwen3.5-27B-AWQ-BF16-INT4 on 2x Multi-GPU (vllm AWQ) | 24 | — | Imported | 2026-05-06 |
| 416 | Devstral-Small-2-24B-Instruct-2512 on 3x AMD Radeon AI Pro R9700 (llama.cpp Q8_0) | 23.87 | — | Imported | 2026-05-06 |
| 417 | Qwen3.5-35B-A3B-GGUF on UNIFIED (llama.cpp UD-IQ4_XS) | 23.56 | — | Imported | 2026-05-06 |
| 418 | MiniMax-M2.7 on UNIFIED (llama.cpp UD-Q3_K_S) | 23.04 | — | Imported | 2026-05-06 |
| 419 | Qwen3.6-27B-TQ3_4S on 5060 Ti (llama.cpp Q3_K_S) | 22.99 | — | Imported | 2026-05-06 |
| 420 | Qwen3-32B on 2x AMD Radeon AI Pro R9700 (vllm FP8) | 22.87 | — | Imported | 2026-05-06 |
| 421 | Qwen3-32B on 3x AMD Radeon AI Pro R9700 (vllm FP8) | 22.76 | — | Imported | 2026-05-06 |
| 422 | Qwen3.6-27B-FP8 on 4x Intel Arc Pro B70 (vllm FP8 compressed-tensors) | 22.72 | — | Imported | 2026-05-06 |
| 423 | Qwen3.6-27B on 2x RTX 3090 (llama.cpp Q6_K_XL) | 22.70 | — | Imported | 2026-05-06 |
| 424 | Qwen3.6-27B-GGUF on 2x RTX 3090 (llama.cpp UD-Q6_K_XL) | 22.23 | — | Imported | 2026-05-06 |
| 425 | Ornstein-3.6-27B-GGUF on UNIFIED (llama.cpp Q4_K_M) | 22 | — | Imported | 2026-05-06 |
| 426 | Qwen3.6-27B on UNIFIED (lmstudio Q4_K_M) | 21.84 | — | Imported | 2026-05-06 |
| 427 | gemma-4-E4B-it-GGUF on UNIFIED (llama.cpp UD-Q4_K_XL) | 21.70 | — | Imported | 2026-05-06 |
| 428 | Qwen3.6-35B-A3B on GTX 1080 Ti (llama.cpp Q2_K) | 21.60 | — | Imported | 2026-05-06 |
| 429 | Qwen3.5-122B-A10B on UNIFIED (llama.cpp UD-IQ4_NL) | 21.25 | — | Imported | 2026-05-06 |
| 430 | Ornstein-27B-v2 on UNIFIED (llama.cpp IQ4_NL) | 20.98 | — | Imported | 2026-05-06 |
| 431 | Kimi-Dev-72B on 3x AMD Radeon AI Pro R9700 (llama.cpp Q4_0) | 20.93 | — | Imported | 2026-05-06 |
| 432 | Qwen3.6-27B-FP8 on 2x Intel Arc Pro B70 32GB (vllm fp8) | 20.11 | — | Imported | 2026-05-06 |
| 433 | Qwen3.6-27B on 3x AMD Radeon AI Pro R9700 (llama.cpp Q8_0) | 20.09 | — | Imported | 2026-05-06 |
| 434 | Qwen3.6-27B on 3x AMD Radeon AI Pro R9700 (llama.cpp Q8_0) | 19.94 | — | Imported | 2026-05-06 |
| 435 | gemma-4-26B-A4B-it-GGUF on UNIFIED (llama.cpp UD-Q4_K_M) | 19.50 | — | Imported | 2026-05-06 |
| 436 | Qwen3.6-35B-A3B on GTX 1080 Ti (llama.cpp Q2_K_XL) | 19.40 | — | Imported | 2026-05-06 |
| 437 | Qwen3.6-27B-GGUF on Intel Arc Pro B70 (llama.cpp Q4_0) | 19.25 | — | Imported | 2026-05-06 |
| 438 | Qwen3.6-27B-GGUF on 2x RTX 3090 (llama.cpp UD-Q8_K_XL) | 19.22 | — | Imported | 2026-05-06 |
| 439 | Qwen3.6-35B-A3B on GTX 1080 Ti (llama.cpp IQ2_M) | 19.10 | — | Imported | 2026-05-06 |
| 440 | Qwen3.5-122B-A10B on UNIFIED (llama.cpp UD-IQ4_NL) | 19.02 | — | Imported | 2026-05-06 |
| 441 | Qwen3.6-27B on UNIFIED (lmstudio Q4_K_M) | 19.01 | — | Imported | 2026-05-06 |
| 442 | Qwen3.6-27B on UNIFIED (lmstudio 8bit-MLX) | 18.52 | — | Imported | 2026-05-06 |
| 443 | Qwen3.6-27B-UD-MLX-4bit on UNIFIED (mlx int4) | 18.43 | — | Imported | 2026-05-06 |
| 444 | Qwen3.6-27B-UD-MLX-4bit on UNIFIED (mlx int4) | 18.43 | — | Imported | 2026-05-06 |
| 445 | Qwen3.6-27B on 3x AMD Radeon AI Pro R9700 (llama.cpp Q8_0) | 18.09 | — | Imported | 2026-05-06 |
| 446 | Qwen3.5-27B on Intel Arc Pro B70 (llama.cpp Q4_K_XL) | 18.02 | — | Imported | 2026-05-06 |
| 447 | Qwen3.6-27B-AEON-Ultimate-Uncensored-NVFP4 on UNIFIED (vllm nvfp4) | 18 | — | Imported | 2026-05-06 |
| 448 | Ornstein-Hermes-3.6-27b-MLX-8bit on UNIFIED (mlx 8bit) | 18 | — | Imported | 2026-05-06 |
| 449 | Qwen3.6-27B on 2x Tesla P100-PCIE-16GB (llama.cpp Q5_K_M) | 17.91 | — | Imported | 2026-05-06 |
| 450 | gemma-4-31B-it-MLX-4bit on UNIFIED (lmstudio 4bit) | 17.84 | — | Imported | 2026-05-06 |
| 451 | MiniMax-M2.5-NVFP4 on UNIFIED (vllm NVFP4) | 17.70 | — | Imported | 2026-05-06 |
| 452 | granite-4.1-30b on 3x AMD Radeon AI Pro R9700 (llama.cpp Q8_0) | 17.28 | — | Imported | 2026-05-06 |
| 453 | Qwen3.5-122B-A10B on UNIFIED (llama.cpp UD-Q6_K_XL) | 17.11 | — | Imported | 2026-05-06 |
| 454 | Qwen3.6-27B on UNIFIED (hipfire MQ4) | 17 | — | Imported | 2026-05-06 |
| 455 | DeepSeek-V4-Flash-2bit-DQ on UNIFIED (mlx 2bit-DQ) | 17 | — | Imported | 2026-05-06 |
| 456 | Qwen3.6-27B on UNIFIED (llama.cpp Q4_K_XL) | 16.90 | — | Imported | 2026-05-06 |
| 457 | Ornstein-27B-v2 on UNIFIED (llama.cpp IQ4_NL) | 16.11 | — | Imported | 2026-05-06 |
| 458 | Ministral-3-3B-Reasoning-2512 on UNIFIED (llama.cpp Q4_K_M) | 15.90 | — | Imported | 2026-05-06 |
| 459 | Qwen3.5-9B on RTX 3080 Ti (llama.cpp Q8_0) | 15.36 | — | Imported | 2026-05-06 |
| 460 | gemma-4-31B-it on 2x Tesla P100-PCIE-16GB (llama.cpp Q5_K_M) | 15.29 | — | Imported | 2026-05-06 |
| 461 | Qwen3.6-35B-A3B on GTX 1060 6GB (llama.cpp UD-IQ2_M) | 15 | — | Imported | 2026-05-06 |
| 462 | Qwen3.5-27B on AMD Radeon 8060S Graphics (Strix Halo APU, gfx1151) (hipfire MQ4) | 14.79 | — | Imported | 2026-05-06 |
| 463 | Qwen3.6-35B-A3B on GTX 1080 Ti (llama.cpp IQ3_S) | 14.60 | — | Imported | 2026-05-06 |
| 464 | gemma-4-31B-it-GGUF on 2x RTX 3090 (llama.cpp UD-Q6_K_XL) | 14.38 | — | Imported | 2026-05-06 |
| 465 | Qwen3.6-27B-FP8 on UNIFIED (vllm fp8) | 14.29 | — | Imported | 2026-05-06 |
| 466 | Qwen3.6-27B on UNIFIED (llama.cpp UD-Q8_K_XL) | 14.25 | — | Imported | 2026-05-06 |
| 467 | Qwen3.6-27B on UNIFIED (llama.cpp UD-Q8_K_XL) | 14.04 | — | Imported | 2026-05-06 |
| 468 | MiniMax-M2.7-GGUF on UNIFIED (llama.cpp UD-Q3_K_M) | 13.95 | — | Imported | 2026-05-06 |
| 469 | Qwen3.6-27B-UD-MLX-6bit on UNIFIED (mlx 6bit) | 13 | — | Imported | 2026-05-06 |
| 470 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q4_K_M) | 12.97 | — | Imported | 2026-05-06 |
| 471 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q4_K_M) | 12.94 | — | Imported | 2026-05-06 |
| 472 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q4_K_M) | 12.86 | — | Imported | 2026-05-06 |
| 473 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q4_K_M) | 12.86 | — | Imported | 2026-05-06 |
| 474 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q4_K_M) | 12.85 | — | Imported | 2026-05-06 |
| 475 | Ornstein-27B-v2 on UNIFIED (llama.cpp IQ4_NL) | 12.83 | — | Imported | 2026-05-06 |
| 476 | Ornstein-27B-v2 on UNIFIED (llama.cpp IQ4_NL) | 12.78 | — | Imported | 2026-05-06 |
| 477 | Qwen3.6-35B-A3B on AMD Radeon RX 9070 XT (llama.cpp UD-Q2_K_XL) | 12.72 | — | Imported | 2026-05-06 |
| 478 | Qwen3.6-27B on UNIFIED (llama.cpp Q4_0) | 12.68 | — | Imported | 2026-05-06 |
| 479 | Qwen3.5-27B on UNIFIED (llama.cpp IQ4_NL) | 12.64 | — | Imported | 2026-05-06 |
| 480 | Ornstein-27B-v2 on UNIFIED (llama.cpp Q4_K_M) | 12.54 | — | Imported | 2026-05-06 |
| 481 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q4_K_M) | 12.52 | — | Imported | 2026-05-06 |
| 482 | Qwen3.6-27B on UNIFIED (llama.cpp Q4_K_M) | 12.49 | — | Imported | 2026-05-06 |
| 483 | Carnice-27b on UNIFIED (llama.cpp Q4_K_M) | 12.46 | — | Imported | 2026-05-06 |
| 484 | Qwen3.6-27B on UNIFIED (llama.cpp Q4_K_M) | 12.45 | — | Imported | 2026-05-06 |
| 485 | Carnice-27b on UNIFIED (llama.cpp Q4_K_M) | 12.44 | — | Imported | 2026-05-06 |
| 486 | Qwen3.6-27B on UNIFIED (llama.cpp Q4_K_XL) | 12.40 | — | Imported | 2026-05-06 |
| 487 | Qwen3.5-27B on UNIFIED (llama.cpp Q4_K_M) | 12.30 | — | Imported | 2026-05-06 |
| 488 | Qwen3.5-27B on UNIFIED (llama.cpp Q4_K_M) | 12.27 | — | Imported | 2026-05-06 |
| 489 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q4_K_M) | 12.19 | — | Imported | 2026-05-06 |
| 490 | Ornstein-27B-v2 on UNIFIED (llama.cpp Q4_K_M) | 12.14 | — | Imported | 2026-05-06 |
| 491 | Qwen3.6-27B on UNIFIED (llama.cpp Q4_K_M) | 12.06 | — | Imported | 2026-05-06 |
| 492 | Qwen3.6-27B on UNIFIED (llama.cpp UD-Q4_K_XL) | 12.04 | — | Imported | 2026-05-06 |
| 493 | Qwen3.6-27B on UNIFIED (llama.cpp UD-Q4_K_XL) | 12.02 | — | Imported | 2026-05-06 |
| 494 | Qwen3.6-27B on UNIFIED (llama.cpp Q4_K_M) | 11.94 | — | Imported | 2026-05-06 |
| 495 | Qwen3.5-27B on UNIFIED (llama.cpp Q4_K_M) | 11.92 | — | Imported | 2026-05-06 |
| 496 | Qwen3.5-27B on UNIFIED (llama.cpp Q4_K_M) | 11.92 | — | Imported | 2026-05-06 |
| 497 | Qwen3.6-27B on UNIFIED (llama.cpp UD-Q4_K_XL) | 11.63 | — | Imported | 2026-05-06 |
| 498 | Qwen3.5-27B on UNIFIED (vllm NVFP4) | 11.49 | — | Imported | 2026-05-06 |
| 499 | Qwen3.6-27B-GGUF on UNIFIED (llama.cpp UD-Q4_K_XL) | 11.46 | — | Imported | 2026-05-06 |
| 500 | Kimi-Dev-72B on 3x AMD Radeon AI Pro R9700 (llama.cpp Q4_0) | 11.35 | — | Imported | 2026-05-06 |
| 501 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q5_K_M) | 11.32 | — | Imported | 2026-05-06 |
| 502 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q5_K_M) | 11.32 | — | Imported | 2026-05-06 |
| 503 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q5_K_M) | 11.27 | — | Imported | 2026-05-06 |
| 504 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q5_K_M) | 11.26 | — | Imported | 2026-05-06 |
| 505 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q5_K_M) | 11.26 | — | Imported | 2026-05-06 |
| 506 | Qwen3.6-35B-A3B on GTX 1080 Ti (llama.cpp Q4_K_M) | 11.10 | — | Imported | 2026-05-06 |
| 507 | Qwen3.5-27B-GGUF on UNIFIED (llama.cpp UD-Q4_K_XL) | 11.10 | — | Imported | 2026-05-06 |
| 508 | gemma-4-31B-it on UNIFIED (llama.cpp Q4_K_M) | 10.69 | — | Imported | 2026-05-06 |
| 509 | gemma-4-31B-it on UNIFIED (llama.cpp Q4_K_M) | 10.69 | — | Imported | 2026-05-06 |
| 510 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q6_K) | 9.86 | — | Imported | 2026-05-06 |
| 511 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q6_K) | 9.85 | — | Imported | 2026-05-06 |
| 512 | Ornstein-27B-v2 on UNIFIED (llama.cpp Q6_K) | 9.70 | — | Imported | 2026-05-06 |
| 513 | Qwen3.6-35B-A3B on GTX 1080 Ti (llama.cpp Q4_K_XL) | 9.50 | — | Imported | 2026-05-06 |
| 514 | Ornstein-27B-v2 on UNIFIED (llama.cpp Q6_K) | 9.44 | — | Imported | 2026-05-06 |
| 515 | Qwen3.5-35B-A3B on GTX 1080 Ti (llama.cpp Q4_K_XL) | 9.30 | — | Imported | 2026-05-06 |
| 516 | Qwen3.6-27B on UNIFIED (ollama Q4_K_XL) | 8.30 | — | Imported | 2026-05-06 |
| 517 | Qwen3.6-27B-AEON-Ultimate-Uncensored-NVFP4 on UNIFIED (vllm NVFP4) | 8 | — | Imported | 2026-05-06 |
| 518 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q8_0) | 7.85 | — | Imported | 2026-05-06 |
| 519 | Qwopus3.5-27B-v3 on UNIFIED (llama.cpp Q8_0) | 7.84 | — | Imported | 2026-05-06 |
| 520 | Qwen3.6-27B-GGUF on UNIFIED (llama.cpp UD-Q6_K_XL) | 7.81 | — | Imported | 2026-05-06 |
| 521 | Qwen3.6-27B on Tesla P100-PCIE-16GB (llama.cpp Q3_K_XL) | 7.53 | — | Imported | 2026-05-06 |
| 522 | Qwen3.6-27B on UNIFIED (llama.cpp Q4_K_XL) | 7 | — | Imported | 2026-05-06 |
| 523 | Qwen3.6-27B on UNIFIED (llama.cpp UD-Q8_K_XL) | 6.44 | — | Imported | 2026-05-06 |
| 524 | Qwen3.6-27B on UNIFIED (llama.cpp UD-Q8_K_XL) | 6.44 | — | Imported | 2026-05-06 |
| 525 | Qwen3.6-35B-A3B on GTX 1080 Ti (llama.cpp Q8_K_XL) | 6.40 | — | Imported | 2026-05-06 |
| 526 | gemma-4-E2B-it on CPU_ONLY (llama.cpp Q4_K_M) | 6.20 | — | Imported | 2026-05-06 |
| 527 | Qwen3.6-27B on UNIFIED (llama.cpp UD-Q8_K_XL) | 6.13 | — | Imported | 2026-05-06 |
| 528 | Qwen3.6-27B on UNIFIED (llama.cpp UD-Q8_K_XL) | 6.13 | — | Imported | 2026-05-06 |
| 529 | Qwen3.6-27B-GGUF on UNIFIED (llama.cpp UD-Q8_K_XL) | 6.07 | — | Imported | 2026-05-06 |
| 530 | Qwen3.6-27B on UNIFIED (llama.cpp UD-Q8_K_XL) | 5.83 | — | Imported | 2026-05-06 |
| 531 | Llama-3.1-Nemotron-70B-Instruct-HF on UNIFIED (llama.cpp Q4_K_M) | 5.18 | — | Imported | 2026-05-06 |
| 532 | Llama-3.1-Nemotron-70B-Instruct-HF on UNIFIED (llama.cpp Q4_K_M) | 5.17 | — | Imported | 2026-05-06 |
| 533 | Qwen3.5-27B-GGUF on UNIFIED (llama.cpp Q4_0) | 5.03 | — | Imported | 2026-05-06 |
| 534 | Qwen3.6-27B-FP8 on UNIFIED (sglang fp8) | 4.72 | — | Imported | 2026-05-06 |
| 535 | gemma-4-31B-it-GGUF on UNIFIED (llama.cpp UD-Q4_K_XL) | 4 | — | Imported | 2026-05-06 |
| 536 | Qwen3.6-27B-GGUF on UNIFIED (llama.cpp BF16) | 3.81 | — | Imported | 2026-05-06 |
| 537 | Qwen3.5-122B-A10B on GTX 1080 Ti (llama.cpp IQ3_S) | 3.20 | — | Imported | 2026-05-06 |
| 538 | Qwen3.5-27B-GGUF on UNIFIED (llama.cpp Q8_0) | 2.82 | — | Imported | 2026-05-06 |
| 539 | Qwen3.5-27B on GTX 1080 Ti (llama.cpp IQ4_NL) | 2.80 | — | Imported | 2026-05-06 |
| 540 | Qwopus3.6-27B-v1-preview on AMD Radeon RX 9070 XT (llama.cpp Q3_K_L) | 2.72 | — | Imported | 2026-05-06 |
| 541 | Qwen3.6-27B on AMD Radeon RX 9070 XT (llama.cpp UD-IQ3_XXS) | 2.31 | — | Imported | 2026-05-06 |
| 542 | Qwen3.5-27B on GTX 1080 Ti (llama.cpp Q4_K_XL) | 2.20 | — | Imported | 2026-05-06 |
| 543 | Qwopus3.6-27B-v1-preview on AMD Radeon RX 9070 XT (llama.cpp Q3_K_L) | 1.03 | — | Imported | 2026-05-06 |
No matching rows.