COLLECTION / RECIPES

Recipes

One artifact × hardware × engine unit. Validated rows are launch-safe; measured speed sweeps attach to each recipe as evidence.

68 records

AMD Radeon RX 7700 XT1 × 12 GBgemma-4-12B-it-qat-q4_0-unquantizedQ4_0vllmreferenceUnknown contextcontext40.4 tok/smeasuredcandidate vllm evidence measured

Apple M1 Max 64GB1 × 64 GBgemma-4-12BBF16llama.cppreference794context14 tok/smeasuredcandidate llama.cpp evidence measured

Apple M3 Ultra 96GB 60-core GPU1 × 96 GBgemma-4-12BQ4_K_Mllama.cppreference255.9Kcontext52 tok/smeasuredcandidate llama.cpp evidence measured

Apple M3 Ultra 96GB 60-core GPU1 × 96 GBgemma-4-12Bq4kmllama.cppreference255.9Kcontext52 tok/smeasuredcandidate llama.cpp evidence measured

Apple M3 Ultra 96GB 60-core GPU1 × 96 GBgemma-4-12b-coder-fable5-composer2.5-v18bitmlxreference64.1Kcontext40.5 tok/smeasuredcandidate mlx evidence measured

Apple M3 Ultra 96GB 60-core GPU1 × 96 GBgemma-4-12b-q4kmunknownllama.cppreference255.9Kcontext52 tok/smeasuredcandidate llama.cpp evidence measured

Apple M3 Ultra 96GB 80-core GPU1 × 96 GBgemma-4-12BQ4_K_Mllama.cppreference255.9Kcontext53.3 tok/smeasuredcandidate llama.cpp evidence measured

Apple M3 Ultra 96GB 80-core GPU1 × 96 GBgemma-4-12Bq4kmllama.cppreference255.9Kcontext53.3 tok/smeasuredcandidate llama.cpp evidence measured

Apple M3 Ultra 96GB 80-core GPU1 × 96 GBgemma-4-12b-coder-fable5-composer2.5-v18bitmlxreference64.1Kcontext39.4 tok/smeasuredcandidate mlx evidence measured

Apple M3 Ultra 96GB 80-core GPU1 × 96 GBgemma-4-12b-q4kmunknownllama.cppreference255.9Kcontext53.3 tok/smeasuredcandidate llama.cpp evidence measured

Apple M4 Max 36GB1 × 36 GBgemma-4-12BQ4_K_Mllama.cppreference255.9Kcontext38.4 tok/smeasuredcandidate llama.cpp evidence measured

Apple M4 Max 36GB1 × 36 GBgemma-4-12Bq4kmllama.cppreference255.9Kcontext38.4 tok/smeasuredcandidate llama.cpp evidence measured

Apple M4 Max 36GB1 × 36 GBgemma-4-12b-coder-fable5-composer2.5-v18bitmlxreference64.1Kcontext25.1 tok/smeasuredcandidate mlx evidence measured

Apple M4 Max 36GB1 × 36 GBgemma-4-12b-q4kmunknownllama.cppreference255.9Kcontext38.4 tok/smeasuredcandidate llama.cpp evidence measured

Apple M4 Max 128GB1 × 128 GBgemma-4-12BQ4_K_Mllama.cppreference255.9Kcontext46.9 tok/smeasuredcandidate llama.cpp evidence measured

Apple M4 Max 128GB1 × 128 GBgemma-4-12Bq4kmllama.cppreference255.9Kcontext46.9 tok/smeasuredcandidate llama.cpp evidence measured

Apple M4 Max 128GB1 × 128 GBgemma-4-12b-coder-fable5-composer2.5-v18bitmlxreference64.1Kcontext31 tok/smeasuredcandidate mlx evidence measured

Apple M4 Max 128GB1 × 128 GBgemma-4-12b-q4kmunknownllama.cppreference255.9Kcontext46.9 tok/smeasuredcandidate llama.cpp evidence measured

Apple M4 Pro 48GB1 × 48 GBgemma-4-12BQ4_K_Mllama.cppreference255.9Kcontext26.8 tok/smeasuredcandidate llama.cpp evidence measured

Apple M4 Pro 48GB1 × 48 GBgemma-4-12Bq4kmllama.cppreference255.9Kcontext26.8 tok/smeasuredcandidate llama.cpp evidence measured

Apple M4 Pro 48GB1 × 48 GBgemma-4-12b-coder-fable5-composer2.5-v18bitmlxreference64.1Kcontext17.3 tok/smeasuredcandidate mlx evidence measured

Apple M4 Pro 48GB1 × 48 GBgemma-4-12b-q4kmunknownllama.cppreference255.9Kcontext26.8 tok/smeasuredcandidate llama.cpp evidence measured

Apple M5 Max 128GB1 × 128 GBgemma-4-12BQ4_K_Mllama.cppreference255.9Kcontext53.9 tok/smeasuredcandidate llama.cpp evidence measured

Apple M5 Max 128GB1 × 128 GBgemma-4-12Bq4kmllama.cppreference255.9Kcontext53.9 tok/smeasuredcandidate llama.cpp evidence measured