Qwen3-27B

14 records across 2 hardware.

hardware	model	quant	backend	tok/s	pp tok/s	W	ttft s	ctx/tg	date	class	flags	source
RTX 5070 Ti Laptop 12 GB	Qwen3-27B 27B	REAP192	llama.cpp	62.5	–	–	–	16384	2026-09-28	community	–	source
RTX 5070 Ti Laptop 12 GB	Qwen3-27B 27B	REAP192	llama.cpp	58.7	–	–	–	16384	2026-09-29	community	–	source
RTX 5070 Ti Laptop 12 GB	Qwen3-27B 27B	REAP192	llama.cpp	55.2	–	–	–	16384	2026-09-29	community	–	source
RTX 5070 Ti Laptop 12 GB	Qwen3-27B 27B	REAP192	llama.cpp	51.2	–	–	–	16384	2026-09-28	community	–	source
RTX 5070 Ti Laptop 12 GB	Qwen3-27B 27B	REAP192	llama.cpp	40.8	–	–	–	60000	2026-09-28	community	–	source
RTX 5070 Ti Laptop 12 GB	Qwen3-27B 27B	REAP192	llama.cpp	40.8	–	–	–	60000	2026-09-29	community	–	source
RTX 5070 Ti Laptop 12 GB	Qwen3-27B 27B	REAP192	llama.cpp	36.7	–	–	–	16384	2026-09-29	community	–	source
RTX 5070 Ti Laptop 12 GB	Qwen3-27B 27B	REAP192	llama.cpp	31.1	–	–	–	16384	2026-09-28	community	–	source
RTX 5070 Ti Laptop 12 GB	Qwen3-27B 27B	REAP192	llama.cpp	30.0	–	–	–	60000	2026-09-28	community	–	source
RTX 5070 Ti Laptop 12 GB	Qwen3-27B 27B	REAP192	llama.cpp	30.0	–	–	–	60000	2026-09-29	community	–	source
RTX 5070 Ti Laptop 12 GB	Qwen3-27B 27B	REAP192	llama.cpp	28.9	–	–	–	16384	2026-09-29	community	–	source
RTX 5070 Ti Laptop 12 GB	Qwen3-27B 27B	REAP192	llama.cpp	28.7	–	–	–	16384	2026-09-28	community	–	source
M5 Pro	Qwen3-27B	4bit	oMLX	11.2	394.4	–	–	1024	2026-10-09	community	–	source
M5 Pro	Qwen3-27B	4bit	oMLX	10.8	419.1	–	–	4096	2026-10-09	community	–	source