Released Feb 24, 2026
Context window unknown
Parameters 122B
Modalities image, text, video
Weights Open
License Apache 2.0
Commercial use Not allowed
Input price —
Output price — Qwen3.5-122B-A10B is a multimodal Mixture-of-Experts model with 122 billion total parameters and 10 billion activated parameters. It combines strong reasoning, coding, long-context, and visual understanding performance with production-friendly efficiency and a native 262K context window.
Benchmark results Service providers 97.0%
96.7%
94.0%
93.4%
93.3%
93.2%
92.8%
92.1%
91.9%
91.4%
91.3%
90.3%
89.8%
88.4%
87.9%
87.4%
87.3%
87.3%
86.7%
86.7%
86.6%
86.2%
85.9%
85.1%
85.1%
83.9%
83.9%
83.9%
82.9%
82.8%
82.2%
82.0%
81.8%
81.6%
80.8%
79.5%
78.9%
78.3%
77.2%
76.9%
76.6%
76.1%
74.7%
74.4%
72.2%
72.0%
70.4%
69.9%
69.3%
68.9%
67.6%
67.3%
67.1%
66.9%
66.4%
63.8%
63.3%
62.6%
62.0%
61.7%
61.5%
60.5%
60.2%
59.0%
58.7%
58.6%
58.0%
53.2%
49.4%
47.5%
44.5%
44.1%
40.2%
39.5%
36.2%
36.2%
33.6%
24.1%
15.4%
12.7%
9.0%
Pricing, uptime, and speed via OpenRouter — updated Jul 17, 2026, 04:19 AM.
Provider Status Input Output Limits Uptime Speed Notes Alibaba available $0.26/Mtok $2.08/Mtok 262K tokens context 66K tokens max output 100.0% 5m 100.0% 987 ms p50 TTFT 82 tok/s p50 SiliconFlow available $0.26/Mtok $2.08/Mtok 262K tokens context 262K tokens max output 100.0% 1,348 ms p50 TTFT 3.0 tok/s p50 fp8 DeepInfra available $0.29/Mtok $2.40/Mtok 262K tokens context 82K tokens max output 100.0% 708 ms p50 TTFT 53 tok/s p50 fp4 AtlasCloud available $0.30/Mtok cache $0.30/Mtok $2.40/Mtok 262K tokens context 66K tokens max output 100.0% 5m 100.0% 1,590 ms p50 TTFT 83 tok/s p50 fp8 Novita available $0.40/Mtok $3.20/Mtok 262K tokens context 66K tokens max output — 994 ms p50 TTFT 78 tok/s p50 bf16
Qwen3.5-122B-A10B