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

DeepSeek-R1

deepseek· 2025-01-20· 164K ctx· mit
open weightstool callreasoningstructured quant: unknown

spread (output)

5.1×

$1.70 – $8.75 · 22 providers

22 providers priced
ProviderInputOutputCache rdBlended
iFlow
relay router
free/—free/——free/—
NanoGPT
relay router
$0.400$1.70$0.200$0.391
SiliconFlow (China)
relay router
$0.500$2.18—$0.653*
SiliconFlow (China)
relay router
$0.500$2.18—$0.653*
SiliconFlow
relay router
$0.500$2.18—$0.653*
D.Run (China)
relay router
$0.550$2.20—$0.700*
Alibaba (China)
relay router
$0.574$2.29—$0.731*
Eden AI
relay router
$0.700$2.40—$0.855*
Hugging Face
relay router
$0.700$2.50—$0.864*
OpenRouter
unknown
$0.700$2.50—$0.864*
NovitaAI
relay router
$0.700$2.50—$0.864*
Kilo Gateway
relay router
$0.700$2.50—$0.864*
Melious
relay router
$0.696$2.78$0.185$0.561
Vercel AI Gateway
unknown
$1.35$5.40—$1.72*
Amazon Bedrock
hyperscaler
$1.35$5.40—$1.72*
Amazon Bedrock
hyperscaler
$1.35$5.40—$1.72*
Azure
hyperscaler
$1.35$5.40—$1.72*
Azure Cognitive Services
unknown
$1.35$5.40—$1.72*
Merge Gateway
unknown
$1.35$5.40—$1.72*
NanoGPT
relay router
$5.00$6.99$2.50$3.59
Together AI
open weights host
$3.00$7.00—$3.36*
Abacus
relay router
$3.00$7.00—$3.36*
IO.NET
relay router
$2.00$8.75$1.00$2.31

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