{"schemaVersion":"hugging-bay.fit-check.v1","generatedAt":"2026-07-10T04:10:24.625Z","artifactId":"hf-model-sachin-sith-mistral-small-4-119b-2603-mlx-4bit","repo":"sachin-sith/Mistral-Small-4-119B-2603-MLX-4bit","webUrl":"https://huggingbay.xyz/artifact/hf-model-sachin-sith-mistral-small-4-119b-2603-mlx-4bit","name":"Mistral-Small-4-119B-2603-MLX-4bit","paramsB":119,"activeB":119,"quant":"f16","weightFile":null,"fileSizeBytes":1,"ctx":8192,"batch":1,"estimate":{"weightsGb":0,"kvGb":3.22,"overheadGb":0.8,"totalGb":4.02},"gpus":[{"id":"t4-16","label":"NVIDIA T4 16GB (free Colab)","vramGb":16,"unified":false,"verdict":"fits","headroomGb":11.98,"totalGb":4.02},{"id":"l4-24","label":"NVIDIA L4 24GB","vramGb":24,"unified":false,"verdict":"fits","headroomGb":19.98,"totalGb":4.02},{"id":"rtx3060-12","label":"RTX 3060 12GB","vramGb":12,"unified":false,"verdict":"fits","headroomGb":7.98,"totalGb":4.02},{"id":"rtx4080-16","label":"RTX 4080 16GB","vramGb":16,"unified":false,"verdict":"fits","headroomGb":11.98,"totalGb":4.02},{"id":"rtx3090-24","label":"RTX 3090 24GB","vramGb":24,"unified":false,"verdict":"fits","headroomGb":19.98,"totalGb":4.02},{"id":"rtx4090-24","label":"RTX 4090 24GB","vramGb":24,"unified":false,"verdict":"fits","headroomGb":19.98,"totalGb":4.02},{"id":"a100-40","label":"A100 40GB","vramGb":40,"unified":false,"verdict":"fits","headroomGb":35.98,"totalGb":4.02},{"id":"a100-80","label":"A100 80GB","vramGb":80,"unified":false,"verdict":"fits","headroomGb":75.98,"totalGb":4.02},{"id":"h100-80","label":"H100 80GB","vramGb":80,"unified":false,"verdict":"fits","headroomGb":75.98,"totalGb":4.02},{"id":"apple-m-16","label":"Apple M-series 16GB (unified)","vramGb":16,"unified":true,"verdict":"fits","headroomGb":11.98,"totalGb":4.02},{"id":"apple-m-32","label":"Apple M-series 32GB (unified)","vramGb":32,"unified":true,"verdict":"fits","headroomGb":27.98,"totalGb":4.02},{"id":"apple-m-64","label":"Apple M-series 64GB (unified)","vramGb":64,"unified":true,"verdict":"fits","headroomGb":59.98,"totalGb":4.02},{"id":"apple-m-128","label":"Apple M-series 128GB (unified)","vramGb":128,"unified":true,"verdict":"fits","headroomGb":123.98,"totalGb":4.02},{"id":"cpu-ram-32","label":"CPU / 32GB system RAM","vramGb":32,"unified":true,"verdict":"fits","headroomGb":27.98,"totalGb":4.02},{"id":"cpu-ram-64","label":"CPU / 64GB system RAM","vramGb":64,"unified":true,"verdict":"fits","headroomGb":59.98,"totalGb":4.02}],"target":{"id":"rtx4090-24","label":"RTX 4090 24GB","vramGb":24,"maxCtx":58880},"whatIf":[{"quant":"q2_k","bytesPerParam":0.33,"weightsGb":39.27,"totalGb":45.63,"verdict":"no","headroomGb":-21.63},{"quant":"q3_k_m","bytesPerParam":0.44,"weightsGb":52.36,"totalGb":59.77,"verdict":"no","headroomGb":-35.77},{"quant":"q4_k_m","bytesPerParam":0.57,"weightsGb":67.83,"totalGb":76.48,"verdict":"no","headroomGb":-52.48},{"quant":"q5_k_m","bytesPerParam":0.69,"weightsGb":82.11,"totalGb":91.9,"verdict":"no","headroomGb":-67.9},{"quant":"q6_k","bytesPerParam":0.82,"weightsGb":97.58,"totalGb":108.61,"verdict":"no","headroomGb":-84.61},{"quant":"q8_0","bytesPerParam":1.06,"weightsGb":126.14,"totalGb":139.45,"verdict":"no","headroomGb":-115.45},{"quant":"f16","bytesPerParam":2,"weightsGb":238,"totalGb":260.26,"verdict":"no","headroomGb":-236.26}],"guidance":["verdict per GPU: fits (<=90% VRAM) / tight (<=100%) / no.","If it does not fit at this precision, check /api/artifacts/{id}/quantizations or the per-quant rows for a smaller mirror.","Estimates: weights + fp16 KV-cache(ctx,batch) + overhead. Real usage varies by runtime."]}