inferenstack — the market read on AI inference · last observation 2026-09-20. See the boards
Price spread

Qwen3 32B

alibaba· 2025-04· 41K ctx· apache-2.0
open weightstool call quant: unknown

spread (output)

11×

$0.250 – $2.80 · 23 providers

23 providers priced
ProviderInputOutputCache rdBlended
iFlow
relay router
free/—free/——free/—
OVHcloud AI Endpoints
relay router
$0.090$0.250—$0.105*
Deep Infra
open weights host
$0.080$0.280—$0.098*
OpenRouter
unknown
$0.080$0.280—$0.098*
Kilo Gateway
relay router
$0.080$0.280—$0.098*
Abacus
relay router
$0.090$0.290—$0.108*
NanoGPT
relay router
$0.100$0.300$0.050$0.086
Cortecs
relay router
$0.089$0.312—$0.109*
Chutes
relay router
$0.104$0.416$0.010$0.073
Jiekou.AIfp8
relay router
$0.100$0.450—$0.132*
NovitaAIfp8
relay router
$0.100$0.450—$0.132*
DigitalOcean
relay router
$0.250$0.550—$0.277*
SiliconFlow (China)
relay router
$0.140$0.570—$0.179*
SiliconFlow
relay router
$0.140$0.570—$0.179*
Helicone
relay router
$0.290$0.590—$0.317*
Hugging Face
relay router
$0.290$0.590—$0.317*
Merge Gateway
unknown
$0.150$0.600—$0.191*
Amazon Bedrock
hyperscaler
$0.150$0.600—$0.191*
Vercel AI Gateway
unknown
$0.160$0.640—$0.204*
LLM Gateway
relay router
$0.360$0.870—$0.406*
DevPass (LLM Gateway)
relay router
$0.360$0.870—$0.406*
Pioneer
relay router
$0.900$0.900$0.900$0.974
Alibaba (China)outlier
relay router
$0.287$1.15—$0.365*
Alibabaoutlier
first party lab
$0.700$2.80—$0.891*

Data: models.dev (MIT), snapshot 2026-09-20. Prices are vendor-published; verify before purchase. Methodology.