{"data":{"accepted_at":null,"id":"gemma-4-26b-a4b-it-qat-4bit-apple-m5-max-128gb-mlxfast-tp1-sweep","measured_at":"2026-08-30T09:12:00Z","metrics":{"concurrency":8,"inference_engine_version":"gemma4-26b-a4b-mlx-v1","latest_point_at":"2026-08-30T09:12:00Z","max_context_tokens":null,"peak_generation_tps":484.5,"peak_prompt_tps":6950.7,"point_count":1},"recipe_id":"gemma-4-26b-a4b-it-qat-4bit-apple-m5-max-128gb-mlxfast-tp1","rows":[{"concurrency":8,"context_tokens":null,"decode_tok_s":484.5,"decode_tok_s_per_stream":60.562,"output_tokens":null,"peak_vram_gb":null,"prefill_tok_s":6950.7,"samples":1,"status":"observed","ttft_ms_p50":null}],"schema_version":"local-ai-registry/v1","source":{"commit":null,"kind":"leaderboard","paths":["fixtures/gemma4_26b_a4b_track.json"],"repository":"https://github.com/Layr-Labs/mlxfast-gemma4-26b-a4b-engine","url":"https://www.yukon.org/mlxfast"},"relationships":{"recipe":{"api":"/api/v1/recipes/gemma-4-26b-a4b-it-qat-4bit-apple-m5-max-128gb-mlxfast-tp1","href":"/recipes/gemma-4-26b-a4b-it-qat-4bit-apple-m5-max-128gb-mlxfast-tp1","id":"gemma-4-26b-a4b-it-qat-4bit-apple-m5-max-128gb-mlxfast-tp1"}}},"meta":{"source":"registry"}}