Artificial Analysis · Intelligence Index · Non-reasoning Benchmarks, as Artificial Analysis ran them 86% GPQA Diamond 20% Humanity's Last Exam 36% Terminal-Bench Hard 84% τ²-Bench 52% IFBench 64% Long context reasoning
Speed 55 tokens/s
First answer 1.6 s thinking included
Per 1M tokens $0.60 in · $3.60 out
Context 262K tokens Text WebDev Vision
Overall 94th Math 75th Expert 78th Polish 79th Multi-turn 82nd Longer queries 85th Hard prompts 91st Instruction following 92nd Coding 93rd Creative writing 100th
LMArena · Text 94th of 416 · 18th of 226 open models
Rated by version Qwen Qwen3.8-27B rates level with it.
3.5 397B A17B #94 3.5 122B A10B #136 3.5 27B #153 3.5 35B A3B #169 3.8 27B #100 Qwen3.5 397B A17B · 1,442 (1,438–1,445) · 94th
Against list price Rating for the price
1340 1360 1380 1400 1420 1440 1460 1480 1500 1520 $0.03 $0.1 $0.3 $1 $3 $10 $30 Qwen3.5 397B A17B Muse Wonder Gemini 3.8 Flash Gemini 3.7 Flash GLM-5.3 DeepSeek V4.1 Flash GLM-5.3-Flash Qwen3.5 397B A17B · 1,442 · $0.88 per 1M tokens ($0.39 in, $2.34 out) · 262K context
1340 1360 1380 1400 1420 1440 1460 1480 1500 1520 $0.03 $0.1 $0.3 $1 $3 $10 $30 Qwen3.5 397B A17B Muse Wonder Gemini 3.8 Flash Gemini 3.7 Flash GLM-5.3 DeepSeek V4.1 Flash GLM-5.3-Flash Qwen3.5 397B A17B · 1,442 · $0.88 per 1M tokens ($0.39 in, $2.34 out) · 262K context
LMArena , CC BY 4.0, 15 October 2026 · prices OpenRouter , list price blended 3:1, read 21 October 2026 · Artificial Analysis , read 21 October 2026