Qwen3.5-27B qwen3.5-27b

qwen · qwen reasoning efficient

#12 overall

params
27.8B
arch
dense
context
256k
license
apache-2.0 open weights
reasoning
yes

# why ranked

Overall: #12 — score 41.2 ● medium — one signal missing, weights renormalized

signalweightinput (0–100)
aa_intelligence 0.50 33.8
arena_elo 0.30
bench_composite 0.20 59.8

missing signals are dropped and the remaining weights renormalized — never imputed.

full methodology

# trend

collecting history — trends appear after a few daily runs.

# benchmarks

benchmarkscoresourcedate
AA Intelligence Index 33.8 Artificial Analysis
Aa-lcr 0.7 / 1 LLM Stats
Ai2d 0.9 / 1 LLM Stats
Androidworld-sr 0.6 / 1 LLM Stats
Babyvision 0.4 / 1 LLM Stats
Bfcl-v4 0.7 / 1 LLM Stats
Browsecomp 0.6 / 1 LLM Stats
Browsecomp-zh 0.6 / 1 LLM Stats
C-eval 0.9 / 1 LLM Stats
Cc-ocr 0.8 / 1 LLM Stats
Charxiv-r 0.8 / 1 LLM Stats
Codeforces 0.8 / 3000 LLM Stats
Countbench 1.0 LLM Stats
Deep-planning 0.2 / 1 LLM Stats
Dynamath 0.9 / 1 LLM Stats
Vlmsareblind 1.0 / 1 LLM Stats

# where to run

providerquantctx$/M in$/M out$/M blendtpsuptime
Alibabaunknown 256k $0.20$1.56$0.54 99.6%
SiliconFlowfp8 256k $0.25$2.00$0.69 98.5%
AtlasCloudfp8 256k $0.27$2.16$0.74 99.6%
Novitabf16 256k $0.30$2.40$0.82 99.7%
Phalaunknown 256k $0.30$2.40$0.82 91.7%
DeepInfrafp8 256k $0.26$2.60$0.84 87.6%

blended = (3·input + output) / 4 per 1M tokens · cheapest first

source aliases
llmstats
qwen3.5-27b
openrouter
qwen/qwen3.5-27b