{ "query": { "task": "object detection", "power": "5" }, "task": "object detection", "size_class": "nano (YOLO-n)", "batch": "1", "mixed": false, "mixed_warning": null, "groups": [ { "size": "nano (YOLO-n)", "batch": "1", "platforms": 16, "rows": 111 }, { "size": "small (YOLO-s)", "batch": "1", "platforms": 8, "rows": 47 }, { "size": "medium (YOLO-m)", "batch": "1", "platforms": 8, "rows": 30 }, { "size": "medium (YOLO-m)", "batch": "8", "platforms": 4, "rows": 22 }, { "size": "large (YOLO-l)", "batch": "1", "platforms": 4, "rows": 14 }, { "size": "xlarge (YOLO-x)", "batch": "1", "platforms": 4, "rows": 14 }, { "size": "nano (YOLO-n)", "batch": "8", "platforms": 3, "rows": 9 }, { "size": "small (YOLO-s)", "batch": "8", "platforms": 3, "rows": 24 }, { "size": "mobile-class CNN", "batch": "1", "platforms": 1, "rows": 1 }, { "size": "mid-size CNN", "batch": "unknown", "platforms": 34, "rows": 125 }, { "size": "medium (YOLO-m)", "batch": "unknown", "platforms": 6, "rows": 11 }, { "size": "mobile-class CNN", "batch": "unknown", "platforms": 6, "rows": 30 }, { "size": "small (YOLO-s)", "batch": "unknown", "platforms": 4, "rows": 8 }, { "size": "unknown", "batch": "1", "platforms": 7, "rows": 78 }, { "size": "unknown", "batch": "unknown", "platforms": 37, "rows": 374 } ], "notes": [], "count": 16, "results": [ { "platform": "Hailo-8 (M.2/PCIe, x86 host)", "platform_url": "/platform/hailo-8-m-2-pcie-x86-host.html", "perf": 1036, "perf_unit": "FPS", "metric": "throughput", "model": "yolov8n", "precision": "INT8 (Hailo quantized)", "stack": "Hailo DFC v2.20.0 / HailoRT (model zoo)", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "D", "confidence_label": "D – vendor-published", "power_w": 2.5, "power_label": "reported in result source — spec", "power_kind": "spec", "perf_per_w": 414.4, "date": "undated", "older_than_18_months": false, "superseded": false, "power_basis": "chip typical (vendor spec)", "price_usd": null, "price_label": "unknown", "form_factor": "AI accelerator (M.2)", "vendor": "Hailo", "source": { "name": "Hailo Model Zoo public models (Hailo-8)", "url": "https://github.com/hailo-ai/hailo_model_zoo/tree/v2.20.0/docs/public_models/HAILO8", "date": null }, "power_source": "https://github.com/hailo-ai/hailo_model_zoo/tree/v2.20.0/docs/public_models/HAILO8", "price_source": null, "fit": { "perf": null, "power": true, "budget": null }, "meets_all": true, "unknowns": 0, "eligible": true, "headroom": null, "rank_score": 1, "unknown_badges": [ "price unknown" ] }, { "platform": "Hailo-10H (M.2/PCIe, x86 host)", "platform_url": "/platform/hailo-10h-m-2-pcie-x86-host.html", "perf": 375, "perf_unit": "FPS", "metric": "throughput", "model": "yolov8n", "precision": "INT8 (Hailo quantized)", "stack": "Hailo DFC v5.4.0 / HailoRT (model zoo)", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "D", "confidence_label": "D – vendor-published", "power_w": 2.5, "power_label": "reported in result source — spec", "power_kind": "spec", "perf_per_w": 150, "date": "undated", "older_than_18_months": false, "superseded": false, "power_basis": "module typical (vendor spec)", "price_usd": null, "price_label": "unknown", "form_factor": "AI accelerator (M.2)", "vendor": "Hailo", "source": { "name": "Hailo Model Zoo public models (Hailo-10H)", "url": "https://github.com/hailo-ai/hailo_model_zoo/tree/v5.4.0/docs/public_models/HAILO10H", "date": null }, "power_source": "https://github.com/hailo-ai/hailo_model_zoo/tree/v5.4.0/docs/public_models/HAILO10H", "price_source": null, "fit": { "perf": null, "power": true, "budget": null }, "meets_all": true, "unknowns": 0, "eligible": true, "headroom": null, "rank_score": 0.362, "unknown_badges": [ "price unknown" ] }, { "platform": "Apple M4 (10-core GPU, 120 GB/s)", "platform_url": "/platform/apple-m4-10-core-gpu-120-gb-s.html", "perf": 357.14, "perf_unit": "FPS", "metric": "latency_e2e", "model": "YOLO26n", "precision": "FP16", "stack": "Ultralytics 8.4.168 Core AI (coreai-torch 0.4.3), macOS 27.0", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "B", "confidence_label": "B – documented third-party method", "power_w": null, "power_label": "no power published", "power_kind": "unknown", "perf_per_w": null, "date": "2026 (year only)", "older_than_18_months": false, "superseded": false, "power_basis": "unknown", "price_usd": null, "price_label": "unknown", "form_factor": "laptop/desktop SoC", "vendor": "Apple", "source": { "name": "Ultralytics docs: Apple Core AI integration benchmark", "url": "https://docs.ultralytics.com/integrations/coreai/", "date": "2026 (page)" }, "power_source": null, "price_source": null, "fit": { "perf": null, "power": "unknown", "budget": null }, "meets_all": true, "unknowns": 1, "eligible": true, "headroom": null, "rank_score": 0.3447, "unknown_badges": [ "power unknown", "perf/W unknown", "price unknown" ] }, { "platform": "Intel Core Ultra 7 258V (Lunar Lake)", "platform_url": "/platform/intel-core-ultra-7-258v-lunar-lake.html", "perf": 300.3, "perf_unit": "FPS", "metric": "inference_time", "model": "YOLO26n", "precision": "FP32", "stack": "Ultralytics + OpenVINO 2026.2.1", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "C", "confidence_label": "C – community/review", "power_w": null, "power_label": "no power published", "power_kind": "unknown", "perf_per_w": null, "date": "2026 (year only)", "older_than_18_months": false, "superseded": false, "power_basis": "unknown", "price_usd": null, "price_label": "unknown", "form_factor": "AI PC", "vendor": "Intel", "source": { "name": "Ultralytics docs: Intel OpenVINO integration benchmark", "url": "https://docs.ultralytics.com/integrations/openvino/", "date": "2026 (page)" }, "power_source": null, "price_source": null, "fit": { "perf": null, "power": "unknown", "budget": null }, "meets_all": true, "unknowns": 1, "eligible": true, "headroom": null, "rank_score": 0.2899, "unknown_badges": [ "power unknown", "perf/W unknown", "price unknown" ] }, { "platform": "NVIDIA Jetson Orin NX 16GB", "platform_url": "/platform/nvidia-jetson-orin-nx-16gb.html", "perf": 286.53, "perf_unit": "FPS", "metric": "inference_time", "model": "YOLO26n", "precision": "INT8", "stack": "Ultralytics TensorRT", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "B", "confidence_label": "B – documented third-party method", "power_w": null, "power_label": "no power published", "power_kind": "unknown", "perf_per_w": null, "date": "undated", "older_than_18_months": false, "superseded": false, "power_basis": "unknown", "price_usd": null, "price_label": "unknown", "form_factor": "edge module", "vendor": "NVIDIA", "source": { "name": "Ultralytics docs benchmark table", "url": "https://docs.ultralytics.com/guides/nvidia-jetson/", "date": null }, "power_source": null, "price_source": null, "fit": { "perf": null, "power": "unknown", "budget": null }, "meets_all": true, "unknowns": 1, "eligible": true, "headroom": null, "rank_score": 0.2766, "unknown_badges": [ "power unknown", "perf/W unknown", "price unknown" ] }, { "platform": "Hailo-8L (M.2/PCIe, x86 host)", "platform_url": "/platform/hailo-8l-m-2-pcie-x86-host.html", "perf": 201, "perf_unit": "FPS", "metric": "throughput", "model": "yolov8n", "precision": "INT8 (Hailo quantized)", "stack": "Hailo DFC v2.20.0 / HailoRT (model zoo)", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "D", "confidence_label": "D – vendor-published", "power_w": 1.5, "power_label": "reported in result source — spec", "power_kind": "spec", "perf_per_w": 134, "date": "undated", "older_than_18_months": false, "superseded": false, "power_basis": "chip typical (vendor spec)", "price_usd": null, "price_label": "unknown", "form_factor": "AI accelerator (M.2)", "vendor": "Hailo", "source": { "name": "Hailo Model Zoo public models (Hailo-8L)", "url": "https://github.com/hailo-ai/hailo_model_zoo/tree/v2.20.0/docs/public_models/HAILO8L", "date": null }, "power_source": "https://github.com/hailo-ai/hailo_model_zoo/tree/v2.20.0/docs/public_models/HAILO8L", "price_source": null, "fit": { "perf": null, "power": true, "budget": null }, "meets_all": true, "unknowns": 0, "eligible": true, "headroom": null, "rank_score": 0.2587, "unknown_badges": [ "price unknown" ] }, { "platform": "Intel Core Ultra X7 358H (Panther Lake)", "platform_url": "/platform/intel-core-ultra-x7-358h-panther-lake.html", "perf": 248.14, "perf_unit": "FPS", "metric": "inference_time", "model": "YOLO26n", "precision": "FP16", "stack": "Ultralytics + OpenVINO 2026.2.1", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "C", "confidence_label": "C – community/review", "power_w": null, "power_label": "no power published", "power_kind": "unknown", "perf_per_w": null, "date": "2026 (year only)", "older_than_18_months": false, "superseded": false, "power_basis": "unknown", "price_usd": null, "price_label": "unknown", "form_factor": "AI PC", "vendor": "Intel", "source": { "name": "Ultralytics docs: Intel OpenVINO integration benchmark", "url": "https://docs.ultralytics.com/integrations/openvino/", "date": "2026 (page)" }, "power_source": null, "price_source": null, "fit": { "perf": null, "power": "unknown", "budget": null }, "meets_all": true, "unknowns": 1, "eligible": true, "headroom": null, "rank_score": 0.2395, "unknown_badges": [ "power unknown", "perf/W unknown", "price unknown" ] }, { "platform": "Intel Core Ultra 7 155H (Meteor Lake)", "platform_url": "/platform/intel-core-ultra-7-155h-meteor-lake.html", "perf": 172.71, "perf_unit": "FPS", "metric": "inference_time", "model": "YOLO26n", "precision": "INT8", "stack": "Ultralytics + OpenVINO 2026.2.1", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "C", "confidence_label": "C – community/review", "power_w": null, "power_label": "no power published", "power_kind": "unknown", "perf_per_w": null, "date": "2026 (year only)", "older_than_18_months": false, "superseded": false, "power_basis": "unknown", "price_usd": null, "price_label": "unknown", "form_factor": "AI PC", "vendor": "Intel", "source": { "name": "Ultralytics docs: Intel OpenVINO integration benchmark", "url": "https://docs.ultralytics.com/integrations/openvino/", "date": "2026 (page)" }, "power_source": null, "price_source": null, "fit": { "perf": null, "power": "unknown", "budget": null }, "meets_all": true, "unknowns": 1, "eligible": true, "headroom": null, "rank_score": 0.1667, "unknown_badges": [ "power unknown", "perf/W unknown", "price unknown" ] }, { "platform": "Microchip PolarFire SoC Video Kit (VectorBlox CoreVectorBlox)", "platform_url": "/platform/microchip-polarfire-soc-video-kit-vectorblox-corevectorblox.html", "perf": 66.21, "perf_unit": "FPS", "metric": "inference_time", "model": "yolov5n_512x288", "precision": "INT8 (VNNX)", "stack": "VectorBlox SDK (VNNX), no compression", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "D", "confidence_label": "D – vendor-published", "power_w": null, "power_label": "no power published", "power_kind": "unknown", "perf_per_w": null, "date": "undated", "older_than_18_months": false, "superseded": false, "power_basis": "unknown", "price_usd": null, "price_label": "unknown", "form_factor": "FPGA SoC (edge)", "vendor": "Microchip", "source": { "name": "Microchip VectorBlox SDK tutorial metrics", "url": "https://github.com/Microchip-Vectorblox/VectorBlox-SDK/blob/master/docs/tutorial_metrics_appendix.md", "date": null }, "power_source": null, "price_source": null, "fit": { "perf": null, "power": "unknown", "budget": null }, "meets_all": true, "unknowns": 1, "eligible": true, "headroom": null, "rank_score": 0.0639, "unknown_badges": [ "power unknown", "perf/W unknown", "price unknown" ] }, { "platform": "Raspberry Pi 5 + AI HAT+ (Hailo-8)", "platform_url": "/platform/raspberry-pi-5-ai-hat-hailo-8.html", "perf": 46.08, "perf_unit": "FPS", "metric": "inference_time", "model": "YOLO26n", "precision": "INT8", "stack": "Ultralytics Hailo HEF / HailoRT", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "B", "confidence_label": "B – documented third-party method", "power_w": null, "power_label": "no power published", "power_kind": "unknown", "perf_per_w": null, "date": "2026 (year only)", "older_than_18_months": false, "superseded": false, "power_basis": "unknown", "price_usd": null, "price_label": "unknown", "form_factor": "SBC + NPU HAT", "vendor": "Raspberry", "source": { "name": "Ultralytics docs: Hailo integration benchmark", "url": "https://docs.ultralytics.com/integrations/hailo/", "date": "2026 (page)" }, "power_source": null, "price_source": null, "fit": { "perf": null, "power": "unknown", "budget": null }, "meets_all": true, "unknowns": 1, "eligible": true, "headroom": null, "rank_score": 0.0445, "unknown_badges": [ "power unknown", "perf/W unknown", "price unknown" ] }, { "platform": "Raspberry Pi 5 (CPU only)", "platform_url": "/platform/raspberry-pi-5-cpu-only.html", "perf": 14.92, "perf_unit": "FPS", "metric": "inference_time", "model": "YOLO26n", "precision": "FP32/default", "stack": "Ultralytics NCNN", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "B", "confidence_label": "B – documented third-party method", "power_w": null, "power_label": "no power published", "power_kind": "unknown", "perf_per_w": null, "date": "undated", "older_than_18_months": false, "superseded": false, "power_basis": "unknown", "price_usd": null, "price_label": "unknown", "form_factor": "edge module", "vendor": "Raspberry", "source": { "name": "Ultralytics docs benchmark table", "url": "https://docs.ultralytics.com/guides/raspberry-pi/", "date": null }, "power_source": null, "price_source": null, "fit": { "perf": null, "power": "unknown", "budget": null }, "meets_all": true, "unknowns": 1, "eligible": true, "headroom": null, "rank_score": 0.0144, "unknown_badges": [ "power unknown", "perf/W unknown", "price unknown" ] }, { "platform": "NVIDIA DGX Spark", "platform_url": "/platform/nvidia-dgx-spark.html", "perf": 990.1, "perf_unit": "FPS", "metric": "inference_time", "model": "YOLO11n", "precision": "FP16", "stack": "Ultralytics TensorRT", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "B", "confidence_label": "B – documented third-party method", "power_w": 153, "power_label": "platform spec envelope (not measured with this result)", "power_kind": "spec", "perf_per_w": 6.471, "date": "undated", "older_than_18_months": false, "superseded": false, "power_basis": "load ~153 W wall (vminstall); price Tom's Hardware", "price_usd": 4699, "price_label": "indicative list price (third-party/vendor)", "form_factor": "AI mini-PC", "vendor": "NVIDIA", "source": { "name": "Ultralytics docs benchmark table", "url": "https://docs.ultralytics.com/guides/nvidia-dgx-spark/", "date": null }, "power_source": "https://www.tomshardware.com/desktops/mini-pcs/amd-challenges-nvidias-dgx-spark-with-usd3-999-ryzen-ai-halo-with-windows-11-support-strix-halo-desktop-undercuts-nvidia-by-usd700-packs-128gb-of-unified-memory", "price_source": "https://www.tomshardware.com/desktops/mini-pcs/amd-challenges-nvidias-dgx-spark-with-usd3-999-ryzen-ai-halo-with-windows-11-support-strix-halo-desktop-undercuts-nvidia-by-usd700-packs-128gb-of-unified-memory", "fit": { "perf": null, "power": false, "budget": null }, "meets_all": false, "unknowns": 0, "eligible": false, "headroom": null, "rank_score": 0.4857, "unknown_badges": [] }, { "platform": "NVIDIA Jetson AGX Thor", "platform_url": "/platform/nvidia-jetson-agx-thor.html", "perf": 719.42, "perf_unit": "FPS", "metric": "inference_time", "model": "YOLO26n", "precision": "FP16", "stack": "Ultralytics TensorRT", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "B", "confidence_label": "B – documented third-party method", "power_w": 130, "power_label": "platform spec envelope (not measured with this result)", "power_kind": "spec", "perf_per_w": 5.534, "date": "undated", "older_than_18_months": false, "superseded": false, "power_basis": "max TDP (OpenNav); price mypcrig", "price_usd": 3499, "price_label": "indicative list price (third-party/vendor)", "form_factor": "edge module", "vendor": "NVIDIA", "source": { "name": "Ultralytics docs benchmark table", "url": "https://docs.ultralytics.com/guides/nvidia-jetson/", "date": null }, "power_source": "https://mypcrig.com/blog/jetson-orin-thor-local-llm-benchmark-2026/", "price_source": "https://mypcrig.com/blog/jetson-orin-thor-local-llm-benchmark-2026/", "fit": { "perf": null, "power": false, "budget": null }, "meets_all": false, "unknowns": 0, "eligible": false, "headroom": null, "rank_score": 0.3539, "unknown_badges": [] }, { "platform": "AMD Ryzen AI Max+ 395 / PRO 395 (Strix Halo, Radeon 8060S)", "platform_url": "/platform/amd-ryzen-ai-max-395-pro-395-strix-halo-radeon-8060s.html", "perf": 476.19, "perf_unit": "FPS", "metric": "inference_time", "model": "YOLO26n", "precision": "FP32", "stack": "Ultralytics 8.4.165 ONNX Runtime MIGraphX EP", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "B", "confidence_label": "B – documented third-party method", "power_w": 120, "power_label": "platform spec envelope (not measured with this result)", "power_kind": "spec", "perf_per_w": 3.968, "date": "2026 (year only)", "older_than_18_months": false, "superseded": false, "power_basis": "spec: SoC cTDP 45–120 W, default TDP 55 W (AMD product pages, identical for 395 and PRO 395); 16 Zen 5 cores, Radeon 8060S 40 CUs, NPU up to 50 TOPS, max memory 128 GB; system power not included; price unknown", "price_usd": null, "price_label": "unknown", "form_factor": "AI mini-PC", "vendor": "AMD", "source": { "name": "Ultralytics docs: AMD ROCm/MIGraphX integration benchmark", "url": "https://docs.ultralytics.com/integrations/amd/", "date": "2026 (page)" }, "power_source": "https://www.amd.com/en/products/processors/laptop/ryzen/ai-300-series/amd-ryzen-ai-max-plus-395.html", "price_source": null, "fit": { "perf": null, "power": false, "budget": null }, "meets_all": false, "unknowns": 0, "eligible": false, "headroom": null, "rank_score": 0.2346, "unknown_badges": [ "price unknown" ] }, { "platform": "NVIDIA Jetson AGX Orin 64GB", "platform_url": "/platform/nvidia-jetson-agx-orin-64gb.html", "perf": 434.78, "perf_unit": "FPS", "metric": "inference_time", "model": "YOLO26n", "precision": "INT8", "stack": "Ultralytics TensorRT", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "B", "confidence_label": "B – documented third-party method", "power_w": 65, "power_label": "platform spec envelope (not measured with this result)", "power_kind": "spec", "perf_per_w": 6.689, "date": "undated", "older_than_18_months": false, "superseded": false, "power_basis": "max TDP per OpenNav report / mypcrig price", "price_usd": 1999, "price_label": "indicative list price (third-party/vendor)", "form_factor": "edge module", "vendor": "NVIDIA", "source": { "name": "Ultralytics docs benchmark table", "url": "https://docs.ultralytics.com/guides/nvidia-jetson/", "date": null }, "power_source": "https://opennav.org/news/opennav-robotics-workload-benchmark/", "price_source": "https://opennav.org/news/opennav-robotics-workload-benchmark/", "fit": { "perf": null, "power": false, "budget": null }, "meets_all": false, "unknowns": 0, "eligible": false, "headroom": null, "rank_score": 0.2179, "unknown_badges": [] }, { "platform": "NVIDIA Jetson Orin Nano Super", "platform_url": "/platform/nvidia-jetson-orin-nano-super.html", "perf": 263.16, "perf_unit": "FPS", "metric": "inference_time", "model": "YOLO26n", "precision": "INT8", "stack": "Ultralytics TensorRT", "size_class": "nano (YOLO-n)", "batch": "1", "like_for_like": true, "differs": [], "confidence": "B", "confidence_label": "B – documented third-party method", "power_w": 25, "power_label": "reported in result source — spec", "power_kind": "spec", "perf_per_w": 10.53, "date": "undated", "older_than_18_months": false, "superseded": false, "power_basis": "module max power mode (spec 7–25 W)", "price_usd": 249, "price_label": "indicative list price (third-party/vendor)", "form_factor": "edge module", "vendor": "NVIDIA", "source": { "name": "Ultralytics docs benchmark table", "url": "https://docs.ultralytics.com/guides/nvidia-jetson/", "date": null }, "power_source": "https://docs.ultralytics.com/guides/nvidia-jetson/", "price_source": "https://www.nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-orin/nano-super-developer-kit/", "fit": { "perf": null, "power": false, "budget": null }, "meets_all": false, "unknowns": 0, "eligible": false, "headroom": null, "rank_score": 0.1397, "unknown_badges": [] } ], "no_data": null, "legend": "Confidence: A audited MLPerf closed, B documented third-party, C community, D vendor-published, E estimate. Unknown is never assumed." }