{ "protocolVersion": "osai-bench/1.3", "clientVersion": "0.10.0", "scoringVersion": "osai-bench-derive/1.3", "createdAt": "2026-08-03T04:22:27.992Z", "qualityOverride": false, "cohortEligible": true, "qualityConditions": [], "runtime": { "name": "ollama", "version": "0.32.5", "endpoint": "loopback", "layerAssignment": null, "offloadPlacement": { "source": "ollama.ps.size_vram", "granularity": "bytes", "residentBytes": 19158655254, "vramResidentBytes": 8889995099, "hostResidentBytes": 10268660155, "vramResidentFraction": 0.464019785373189 }, "kvCacheMetadata": { "source": "ollama.show.model_info", "architecture": "cohere2moe", "blockCount": 49, "kvHeadCount": 4, "attentionHeadCount": 32, "embeddingLength": 2048, "headDimension": 128, "resolvedElementType": null, "projectedBytes": null, "calculationAvailable": false, "missingInputs": [ "resolvedKvCacheElementType" ] }, "environment": { "source": "client-process-env", "authoritative": false, "note": "Read from the benchmark client's own process environment. The Ollama server may have been started with a different environment; Ollama exposes no endpoint reporting its resolved configuration.", "declared": { "OLLAMA_KV_CACHE_TYPE": null, "OLLAMA_FLASH_ATTENTION": null, "OLLAMA_CONTEXT_LENGTH": null, "OLLAMA_NUM_PARALLEL": null, "OLLAMA_GPU_OVERHEAD": null, "OLLAMA_SCHED_SPREAD": null, "OLLAMA_LLM_LIBRARY": null, "LLAMA_ARG_FIT": null, "LLAMA_ARG_FIT_TARGET": null, "OLLAMA_MAX_LOADED_MODELS": null, "OLLAMA_KEEP_ALIVE": null, "OLLAMA_DEBUG": null, "OLLAMA_MODELS": false, "OLLAMA_HOST": false }, "declaredNonDefault": [] } }, "model": { "identifier": "north-mini-code-1.0:latest", "digest": "d8b269ad5c7c7144ce104b83ce93bc3efb85e0f74e01be6be5f5d6f7ca90b60f", "family": "cohere2moe", "parameterSize": "30.5B", "quantization": "Q4_K_M", "weightsBytes": 18593967008, "weightsSource": "ollama.tags.size" }, "system": { "cpu": { "model": "12th Gen Intel(R) Core(TM) i9-12900K" }, "gpu": { "present": true, "model": "NVIDIA GeForce RTX 3080", "totalVramBytes": 10737418240, "freeVramBytesAtCheck": 10326376448, "freeVramBytesAtLoad": 1066401792, "utilizationPercentAtCheck": 0, "driverVersion": "595.71.05", "provider": "nvidia-smi" }, "memory": { "totalBytes": 32817025024 }, "os": { "platform": "linux", "version": "Ubuntu 26.04 LTS", "architecture": "x64" }, "power": { "present": false, "onBattery": false } }, "configuration": { "memoryBandwidthGBps": 760, "memoryBandwidthSource": "manufacturer-table", "memoryBandwidthTableVersion": "osai-gpu-memory-bandwidth/2", "memoryBandwidthEntryId": "nvidia-geforce-rtx-3080-10gb", "fixedOptions": { "temperature": 0, "seed": 42, "stream": true }, "workloads": { "w1": { "numPredict": 1, "numCtx": 512, "keepAlive": "5m", "promptTokenRange": null, "warmups": 0, "repetitions": 1, "varyPromptPerCall": false }, "w2": { "numPredict": 128, "numCtx": 4096, "keepAlive": "5m", "promptTokenRange": { "min": 20, "max": 64 }, "warmups": 1, "repetitions": 5, "varyPromptPerCall": false }, "w3": { "numPredict": 1, "numCtx": 4096, "keepAlive": "5m", "promptTokenRange": { "min": 2000, "max": 4095 }, "warmups": 1, "repetitions": 5, "varyPromptPerCall": true }, "w4": { "numPredict": 512, "numCtx": 4096, "keepAlive": "5m", "promptTokenRange": { "min": 20, "max": 64 }, "warmups": 1, "repetitions": 5, "varyPromptPerCall": false } }, "resolved": { "fixedOptions": { "temperature": 0, "seed": 42, "stream": true }, "workloads": { "w1": { "numPredict": 1, "numCtx": 512, "keepAlive": "5m" }, "w2": { "numPredict": 128, "numCtx": 4096, "keepAlive": "5m" }, "w3": { "numPredict": 1, "numCtx": 4096, "keepAlive": "5m" }, "w4": { "numPredict": 512, "numCtx": 4096, "keepAlive": "5m" } }, "runtimeModelParameters": "temperature 1\ntop_p 0.95", "unresolvedRuntimeDefaults": true } }, "rawMeasurements": { "workloads": { "w1": { "warmup": null, "measuredPasses": [ { "index": 1, "valid": true, "attempts": [ { "measurement": { "total_duration": 7367003999, "load_duration": 6997661812, "prompt_eval_count": 30, "prompt_eval_duration": 363635000, "eval_count": 1, "eval_duration": 1000, "timeToFirstTokenMs": 7369.876024999998 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 7367003999, "load_duration": 6997661812, "prompt_eval_count": 30, "prompt_eval_duration": 363635000, "eval_count": 1, "eval_duration": 1000, "timeToFirstTokenMs": 7369.876024999998 }, "validity": { "valid": true, "reasons": [] } } ], "failed": false }, "w2": { "warmup": { "total_duration": 11458242896, "load_duration": 6169999285, "prompt_eval_count": 60, "prompt_eval_duration": 1097736000, "eval_count": 128, "eval_duration": 4186179000, "timeToFirstTokenMs": 7273.873074000003 }, "measuredPasses": [ { "index": 1, "valid": true, "attempts": [ { "measurement": { "total_duration": 4538477943, "load_duration": 219410195, "prompt_eval_count": 60, "prompt_eval_duration": 108107000, "eval_count": 128, "eval_duration": 4196377999, "timeToFirstTokenMs": 343.81217099999776 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 4538477943, "load_duration": 219410195, "prompt_eval_count": 60, "prompt_eval_duration": 108107000, "eval_count": 128, "eval_duration": 4196377999, "timeToFirstTokenMs": 343.81217099999776 }, "validity": { "valid": true, "reasons": [] } }, { "index": 2, "valid": true, "attempts": [ { "measurement": { "total_duration": 4520364782, "load_duration": 210631082, "prompt_eval_count": 60, "prompt_eval_duration": 106468000, "eval_count": 128, "eval_duration": 4186832000, "timeToFirstTokenMs": 334.0735329999952 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 4520364782, "load_duration": 210631082, "prompt_eval_count": 60, "prompt_eval_duration": 106468000, "eval_count": 128, "eval_duration": 4186832000, "timeToFirstTokenMs": 334.0735329999952 }, "validity": { "valid": true, "reasons": [] } }, { "index": 3, "valid": true, "attempts": [ { "measurement": { "total_duration": 4521429503, "load_duration": 210918174, "prompt_eval_count": 60, "prompt_eval_duration": 116915000, "eval_count": 128, "eval_duration": 4189198000, "timeToFirstTokenMs": 332.3522710000034 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 4521429503, "load_duration": 210918174, "prompt_eval_count": 60, "prompt_eval_duration": 116915000, "eval_count": 128, "eval_duration": 4189198000, "timeToFirstTokenMs": 332.3522710000034 }, "validity": { "valid": true, "reasons": [] } }, { "index": 4, "valid": true, "attempts": [ { "measurement": { "total_duration": 4535630441, "load_duration": 225349773, "prompt_eval_count": 60, "prompt_eval_duration": 121652000, "eval_count": 128, "eval_duration": 4184721999, "timeToFirstTokenMs": 351.3923619999987 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 4535630441, "load_duration": 225349773, "prompt_eval_count": 60, "prompt_eval_duration": 121652000, "eval_count": 128, "eval_duration": 4184721999, "timeToFirstTokenMs": 351.3923619999987 }, "validity": { "valid": true, "reasons": [] } }, { "index": 5, "valid": true, "attempts": [ { "measurement": { "total_duration": 4520951877, "load_duration": 219501357, "prompt_eval_count": 60, "prompt_eval_duration": 110643000, "eval_count": 128, "eval_duration": 4185212000, "timeToFirstTokenMs": 336.2246670000022 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 4520951877, "load_duration": 219501357, "prompt_eval_count": 60, "prompt_eval_duration": 110643000, "eval_count": 128, "eval_duration": 4185212000, "timeToFirstTokenMs": 336.2246670000022 }, "validity": { "valid": true, "reasons": [] } } ], "failed": false }, "w3": { "warmup": { "total_duration": 6830684821, "load_duration": 202701887, "prompt_eval_count": 2439, "prompt_eval_duration": 6589943000, "eval_count": 1, "eval_duration": 1000, "timeToFirstTokenMs": 6830.989015999992 }, "measuredPasses": [ { "index": 1, "valid": true, "attempts": [ { "measurement": { "total_duration": 7084254233, "load_duration": 212558324, "prompt_eval_count": 2439, "prompt_eval_duration": 6636727000, "eval_count": 1, "eval_duration": 1000, "timeToFirstTokenMs": 7084.671040000001 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 7084254233, "load_duration": 212558324, "prompt_eval_count": 2439, "prompt_eval_duration": 6636727000, "eval_count": 1, "eval_duration": 1000, "timeToFirstTokenMs": 7084.671040000001 }, "validity": { "valid": true, "reasons": [] } }, { "index": 2, "valid": true, "attempts": [ { "measurement": { "total_duration": 7100608477, "load_duration": 242129929, "prompt_eval_count": 2439, "prompt_eval_duration": 6632728000, "eval_count": 1, "eval_duration": 1000, "timeToFirstTokenMs": 7101.035499000005 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 7100608477, "load_duration": 242129929, "prompt_eval_count": 2439, "prompt_eval_duration": 6632728000, "eval_count": 1, "eval_duration": 1000, "timeToFirstTokenMs": 7101.035499000005 }, "validity": { "valid": true, "reasons": [] } }, { "index": 3, "valid": true, "attempts": [ { "measurement": { "total_duration": 7035720535, "load_duration": 237745864, "prompt_eval_count": 2439, "prompt_eval_duration": 6571729000, "eval_count": 1, "eval_duration": 1000, "timeToFirstTokenMs": 7036.106985999999 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 7035720535, "load_duration": 237745864, "prompt_eval_count": 2439, "prompt_eval_duration": 6571729000, "eval_count": 1, "eval_duration": 1000, "timeToFirstTokenMs": 7036.106985999999 }, "validity": { "valid": true, "reasons": [] } }, { "index": 4, "valid": true, "attempts": [ { "measurement": { "total_duration": 6991309810, "load_duration": 204437686, "prompt_eval_count": 2439, "prompt_eval_duration": 6572296000, "eval_count": 1, "eval_duration": 1000, "timeToFirstTokenMs": 6991.590836999996 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 6991309810, "load_duration": 204437686, "prompt_eval_count": 2439, "prompt_eval_duration": 6572296000, "eval_count": 1, "eval_duration": 1000, "timeToFirstTokenMs": 6991.590836999996 }, "validity": { "valid": true, "reasons": [] } }, { "index": 5, "valid": true, "attempts": [ { "measurement": { "total_duration": 7062643317, "load_duration": 210975312, "prompt_eval_count": 2439, "prompt_eval_duration": 6630665000, "eval_count": 1, "eval_duration": 1000, "timeToFirstTokenMs": 7063.012544000012 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 7062643317, "load_duration": 210975312, "prompt_eval_count": 2439, "prompt_eval_duration": 6630665000, "eval_count": 1, "eval_duration": 1000, "timeToFirstTokenMs": 7063.012544000012 }, "validity": { "valid": true, "reasons": [] } } ], "failed": false }, "w4": { "warmup": { "total_duration": 17345765975, "load_duration": 212871985, "prompt_eval_count": 60, "prompt_eval_duration": 105948000, "eval_count": 512, "eval_duration": 16813990000, "timeToFirstTokenMs": 532.1051339999976 }, "measuredPasses": [ { "index": 1, "valid": true, "attempts": [ { "measurement": { "total_duration": 17178879033, "load_duration": 202631913, "prompt_eval_count": 60, "prompt_eval_duration": 107667000, "eval_count": 512, "eval_duration": 16836991000, "timeToFirstTokenMs": 342.20763699998497 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 17178879033, "load_duration": 202631913, "prompt_eval_count": 60, "prompt_eval_duration": 107667000, "eval_count": 512, "eval_duration": 16836991000, "timeToFirstTokenMs": 342.20763699998497 }, "validity": { "valid": true, "reasons": [] } }, { "index": 2, "valid": true, "attempts": [ { "measurement": { "total_duration": 17175986642, "load_duration": 225758695, "prompt_eval_count": 60, "prompt_eval_duration": 114076000, "eval_count": 512, "eval_duration": 16832476999, "timeToFirstTokenMs": 343.83138999997755 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 17175986642, "load_duration": 225758695, "prompt_eval_count": 60, "prompt_eval_duration": 114076000, "eval_count": 512, "eval_duration": 16832476999, "timeToFirstTokenMs": 343.83138999997755 }, "validity": { "valid": true, "reasons": [] } }, { "index": 3, "valid": true, "attempts": [ { "measurement": { "total_duration": 17175084208, "load_duration": 218730702, "prompt_eval_count": 60, "prompt_eval_duration": 113043000, "eval_count": 512, "eval_duration": 16839693000, "timeToFirstTokenMs": 335.87251400001696 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 17175084208, "load_duration": 218730702, "prompt_eval_count": 60, "prompt_eval_duration": 113043000, "eval_count": 512, "eval_duration": 16839693000, "timeToFirstTokenMs": 335.87251400001696 }, "validity": { "valid": true, "reasons": [] } }, { "index": 4, "valid": true, "attempts": [ { "measurement": { "total_duration": 17158732799, "load_duration": 211247296, "prompt_eval_count": 60, "prompt_eval_duration": 111307000, "eval_count": 512, "eval_duration": 16832249000, "timeToFirstTokenMs": 326.9704989999882 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 17158732799, "load_duration": 211247296, "prompt_eval_count": 60, "prompt_eval_duration": 111307000, "eval_count": 512, "eval_duration": 16832249000, "timeToFirstTokenMs": 326.9704989999882 }, "validity": { "valid": true, "reasons": [] } }, { "index": 5, "valid": true, "attempts": [ { "measurement": { "total_duration": 17167500315, "load_duration": 210583415, "prompt_eval_count": 60, "prompt_eval_duration": 110053000, "eval_count": 512, "eval_duration": 16842517000, "timeToFirstTokenMs": 325.26281799998833 }, "validity": { "valid": true, "reasons": [] } } ], "measurement": { "total_duration": 17167500315, "load_duration": 210583415, "prompt_eval_count": 60, "prompt_eval_duration": 110053000, "eval_count": 512, "eval_duration": 16842517000, "timeToFirstTokenMs": 325.26281799998833 }, "validity": { "valid": true, "reasons": [] } } ], "failed": false } } }, "derived": { "generationTokensPerSecond": { "median": 30.409234049005548, "coefficientOfVariation": 0.0002664195828838824, "samples": 5 }, "prefillTokensPerSecond": { "median": 367.8364085653551, "coefficientOfVariation": 0.005105402288339889, "samples": 5 }, "timeToFirstTokenMs": { "median": 336.2246670000022, "coefficientOfVariation": 0.02334011797643363, "samples": 5 }, "coldLoad": { "seconds": 6.997661812, "samples": 1 }, "passFailureRate": { "failedMeasuredPasses": 0, "totalMeasuredPasses": 16, "percent": 0 }, "attemptFailureRate": { "failedAttempts": 0, "totalAttempts": 16, "percent": 0 }, "roofline": { "modelWeightsGB": 18.593967008, "memoryBandwidthGBps": 760, "theoreticalMaxTokensPerSecond": 40.87347254477822, "utilization": 0.7439845982181045 }, "diagnostics": [ { "id": "partial-cpu-offload", "status": "detected", "message": "54% of the model's resident bytes are on the host, not in VRAM", "evidence": { "source": "ollama.ps.size_vram", "granularity": "bytes", "residentBytes": 19158655254, "vramResidentBytes": 8889995099, "hostResidentBytes": 10268660155, "vramResidentFraction": 0.464019785373189 } }, { "id": "context-vram-headroom", "status": "unavailable", "message": "Ollama /api/show does not expose the resolved KV-cache element type actually in use, so v1.1 does not calculate or threshold projected KV-cache VRAM", "evidence": { "configuredContext": 4096, "availableArchitectureMetadata": { "source": "ollama.show.model_info", "architecture": "cohere2moe", "blockCount": 49, "kvHeadCount": 4, "attentionHeadCount": 32, "embeddingLength": 2048, "headDimension": 128, "resolvedElementType": null, "projectedBytes": null, "calculationAvailable": false, "missingInputs": [ "resolvedKvCacheElementType" ] }, "missingInput": "resolvedKvCacheElementType" } }, { "id": "cpu-only-with-gpu", "status": "not-detected", "message": "46% of the model's resident bytes are in VRAM", "evidence": { "source": "ollama.ps.size_vram", "granularity": "bytes", "residentBytes": 19158655254, "vramResidentBytes": 8889995099, "hostResidentBytes": 10268660155, "vramResidentFraction": 0.464019785373189, "gpuModel": "NVIDIA GeForce RTX 3080" } }, { "id": "weights-exceed-vram", "status": "detected", "message": "Quantized on-disk model weight size is larger than total VRAM", "evidence": { "weightsBytes": 18593967008, "totalVramBytes": 10737418240 } } ] } }