COLLECTION / RECIPES

Recipes

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

2,330 records

Ryzen AI Max+ 3951 × 128 GBDeepSeek-V4-Flash-0731ROCMFPX_STRIX_LEAN_2.58BPWluceboxreferenceUnknown contextcontext37.5 tok/smeasuredcandidate lucebox evidence measured

Ryzen AI Max+ 3951 × 128 GBGemma-4-12B-itQ4_K_Mllama.cppreferenceUnknown contextcontext24.8 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBGemma-4-12B-itQ4_K_Mllama.cppreferenceUnknown contextcontext24.7 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBGemma-4-12B-itQ4_K_Mllama.cppreferenceUnknown contextcontext24.8 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBgemma-4-26B-A4BQ4_K_Mllama.cppreferenceUnknown contextcontext65.9 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBgemma-4-26B-A4BQ6_Kllama.cppreferenceUnknown contextcontext55.5 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBgemma-4-26B-A4BQ8_0llama.cppreferenceUnknown contextcontext47.2 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBgemma-4-26B-A4B-itQ4_K_Mllama.cppreferenceUnknown contextcontext66.1 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBgemma-4-31BQ4_K_Mllama.cppreferenceUnknown contextcontext10.7 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBgemma-4-E2BQ4_K_Mllama.cppreferenceUnknown contextcontext107.1 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBgemma-4-E4BQ4_K_Mllama.cppreferenceUnknown contextcontext57.7 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBGemopus-4-26B-A4B-itQ4_K_Mllama.cppreferenceUnknown contextcontext64.3 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBGemopus-4-26B-A4B-itQ8_0llama.cppreferenceUnknown contextcontext46 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBGLM-4.5-AirAWQvllmreference32Kcontext1.1 tok/smeasuredcandidate vllm evidence measured

Ryzen AI Max+ 3951 × 128 GBGLM-4.7-FlashQ4_K_XLllama.cppreferenceUnknown contextcontext53 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBgpt-oss-120bF16llama.cppreference128Kcontext38.8 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBKAT-Coder-V2.5-DevQ4_0_ROCMFP4llama.cppreference256Kcontext66.3 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBKAT-Coder-V2.5-DevQ4_0_ROCMFP4_STRIX_LEANllama.cppreference256Kcontext74.4 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBlaguna-s-2.1Q4_0_ROCMFP4_FASTllama.cppreference64Kcontext40.7 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBlaguna-s-2.1Q4_0_ROCMFP4_STRIX_LEANllama.cppreference256Kcontext41.7 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBlaguna-s-2.1UD-Q4_K_XLllama.cppreference64Kcontext29.9 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBLing-3.0-flashQ4_0_ROCMFP4_STRIXllama.cppreferenceUnknown contextcontext39.2 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBLlama-3.1-8BQ5_K_Mllama.cppreferenceUnknown contextcontext38.6 tok/smeasuredcandidate llama.cpp evidence measured

Ryzen AI Max+ 3951 × 128 GBllama-3.1-70b-instructQ4_K_Mllama.cppreferenceUnknown contextcontext5.2 tok/smeasuredcandidate llama.cpp evidence measured