Qwen3 14B qwen3-14b

qwen · qwen

unranked — no quality signals yet

params
14.8B
arch
dense
context
40k
license
apache-2.0 open weights
released
Apr 2025
downloads/30d
10.4M

# architecture

attn shared GQA 40q/8kv feed-forward (dense) every parameter, every token out ×40 layers ctx 40,960 14.8B — no routing
layers
40
d_model
5120
heads
40q / 8kv
head dim
128
vocab
151936
family
qwen3

signal path of one layer, generated from the registry's structured fields — dimension lines quote the model's real numbers; an MoE trace forks at the router, a dense trace runs straight through. Geometry fields come from the repo's config.json.

# why ranked

Overall: provisional — score 19.0 ● low — a single signal; treat with caution

provisional: too few independent signals for a numbered position — the score is shown, but this model sorts after every ranked model.

signalweightinput (0–100)
aa_intelligence 0.70 —
bench_composite 0.30 19.0

benchmark panel evidence:

complete panelscore (0–100)
aime-2025-v119.0

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

full methodology

# trend

OGM score · last 101 days
OGM score over 101 days

methodology changed during this history window; score movement across that boundary is not model movement. See methodology v6.

# benchmarks

benchmarkscoresourcedate
AIME 2024 0.8 / 1 LLM Stats —
AIME 2025 0.7 / 1 LLM Stats —
Alignbench 0.9 / 1 LLM Stats —
Arena-hard 0.9 / 1 LLM Stats —
Autologi 0.9 / 1 LLM Stats —
Bfcl-v3 0.7 / 1 LLM Stats —
C-eval 0.9 / 1 LLM Stats —
Codeforces 1766.0 / 3000 LLM Stats —
Math-500 1.0 / 1 LLM Stats —

# where to run

providerquantctx$/M in$/M out$/M cacheprice srctpsuptime✓
Featherless — ——— via hfrouter — —
SiliconFlow — ——— siliconflow — — ✓
Nscale 40k $0.07$0.20— via hfrouter 90 —
Nebius 32k $0.08$0.24— via litellm — —
NextBitint4 40k $0.10$0.22— via openrouter — 99.9%
DeepInfrafp8 40k $0.12$0.24— deepinfra — 100.0% ✓
Alibaba 128k $0.35$1.40— via modelsdev — —

sorted by blended price ((3·input + output) / 4 per 1M) · ✓ = the provider's own catalog confirms the offer · "via …" prices are what the aggregator routing the offer charges, not the provider's own list price

source aliases
openrouter
qwen/qwen3-14b