Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input.
C++
[
{
"model": "CODER",
"tier": "1k",
"prompt_tokens": 1017,
"exit": 0,
"wall_s": 18.0,
"prefill_tok_s": 599.05,
"ttft_s": 2.54,
"decode_tok_s": 53.33,
"decoded": 256,
"spec_accept": 0.72,
"tokens_per_round": 2.86,
"vram_slots": 2570
},
{
"model": "CODER",
"tier": "4k",
"prompt_tokens": 3562,
"exit": 0,
"wall_s": 19.6,
"prefill_tok_s": 1152.17,
"ttft_s": 3.94,
"decode_tok_s": 50.6,
"decoded": 256,
"spec_accept": 0.707,
"tokens_per_round": 2.88,
"vram_slots": 2615
},
{
"model": "CODER",
"tier": "32k",
"prompt_tokens": 31566,
"exit": 0,
"wall_s": 40.7,
"prefill_tok_s": 1297.8,
"ttft_s": 25.17,
"decode_tok_s": 53.34,
"decoded": 256,
"spec_accept": 0.785,
"tokens_per_round": 3.02,
"vram_slots": 2418
},
{
"model": "CODER",
"tier": "64k",
"prompt_tokens": 64162,
"exit": 0,
"wall_s": 75.5,
"prefill_tok_s": 1350.09,
"ttft_s": 49.01,
"decode_tok_s": 50.84,
"decoded": 256,
"spec_accept": 0.696,
"tokens_per_round": 2.61,
"vram_slots": 2382
},
{
"model": "CODER",
"tier": "128k",
"prompt_tokens": 128478,
"exit": 0,
"wall_s": 121.5,
"prefill_tok_s": 1266.31,
"ttft_s": 102.95,
"decode_tok_s": 44.03,
"decoded": 256,
"spec_accept": 0.619,
"tokens_per_round": 2.33,
"vram_slots": 2326
},
{
"model": "CODER",
"tier": "262k",
"prompt_tokens": 259943,
"exit": 0,
"wall_s": 280.7,
"prefill_tok_s": 1033.91,
"ttft_s": 252.92,
"decode_tok_s": 42.8,
"decoded": 256,
"spec_accept": 0.684,
"tokens_per_round": 2.6,
"vram_slots": 2208
}
]