DeepSeek V4.1 Flash
deepseek-ai/deepseek-v4.1-flash
spread (output, p90 ÷ p10)
2.1×
$0.280 – $1.50 · 14 providers · max÷min 5.4×
Cheapest / median / spread use providers that serve the model on their own capacity: first-party lab, hyperscaler, inference host.
blended = 90.9% input × [(1 − 70% cache) × input + 70% cache × cache-read] + 9.1% × output + cache writes (30% of uncached input) where a write price is published
A provider that publishes no cache price is blended with cache share 0 (all input at the input price) and labelled “no cache price published” — never penalised against identical in/out prices.
| Provider | Input | Output | Cache rd | Blended |
|---|---|---|---|---|
AMDoutlier inference host | $0.140 | $0.280 | $0.00280 | $0.0654 |
NanoGPT router / aggregator | $0.100 | $0.400 | $0.00300 | $0.0655 |
OpenRouter router / aggregator | $0.150 | $0.600 | $0.00300 | $0.0974 |
Eden AI router / aggregator | $0.150 | $0.600 | $0.0150 | $0.105 |
Merge Gateway router / aggregator | $0.150 | $0.600 | $0.00300 | $0.0974 |
LLM Gateway router / aggregator | $0.150 | $0.600 | $0.0100 | $0.102 |
OpenCode Go subscription plan | $0.150 | $0.600 | $0.00300 | $0.0974 |
Ollama Cloud inference host | $0.150 | $0.600 | $0.00300 | $0.0974 |
Deep Infra inference host | $0.200 | $0.600 | $0.00600 | $0.113 |
AIHubMix router / aggregator | $0.155 | $0.620 | $0.00310 | $0.101 |
Fireworks AI inference host | $0.220 | $0.660 | $0.00700 | $0.124 |
Ofox reseller | $0.210 | $0.840 | $0.00420 | $0.136 |
Vancine reseller | $0.240 | $0.960 | $0.00480 | $0.156 |
CrossModel router / aggregator | $0.27 | $1.08 | $0.00540 | $0.20 |
Melious unclassified | $0.23 | $1.16 | $0.01 | $0.18 |
Alibaba (China)CN first-party lab | $0.30 | $1.19 | $0.01 | $0.20 |
Kilo Gateway router / aggregator | $0.30 | $1.20 | $0.00600 | $0.19 |
Charm Hyper unclassified | $0.30 | $1.20 | $0.03 | $0.21 |
Hugging Face inference host | $0.30 | $1.20 | — | $0.38no cache price published |
Vercel AI Gateway router / aggregator | $0.30 | $1.20 | $0.03 | $0.21 |
EmpirioLabs AI unclassified | $0.30 | $1.20 | $0.30 | $0.38 |
Nebius Token Factory inference host | $0.30 | $1.20 | $0.30 | $0.38 |
OpenCode Zen router / aggregator | $0.30 | $1.20 | $0.00600 | $0.19 |
DigitalOcean inference host | $0.30 | $1.20 | $0.00600 | $0.19 |
Baseten inference host | $0.30 | $1.20 | $0.03 | $0.21 |
Together AI inference host | $0.30 | $1.20 | $0.00600 | $0.19 |
Synthetic inference host | $0.60 | $1.20 | $0.03 | $0.29 |
GreenPT inference host | $0.26 | $1.28 | $0.01 | $0.19 |
Tinfoil inference host | $0.65 | $1.45 | $0.13 | $0.39 |
Cortecs router / aggregator | $0.50 | $1.50 | $0.13 | $0.36 |
Venice AI inference host | $0.38 | $1.50 | $0.00750 | $0.24 |
TensorX unclassified | $0.50 | $1.50 | $0.13 | $0.36 |
Requesty router / aggregator | $0.50 | $1.50 | $0.05 | $0.30 |
Inco unclassified | $0.60 | $2.40 | — | $0.76no cache price published |
Subscription plans (flat fee, not per-token) · 5
Flat-fee plan — the per-token price is bundled into a subscription, so $0/token is not free. Excluded from cheapest, median and spread.
- SCNet Token Plan DeepSeek-V4.1-FlashMainland-China endpointsubscription plan
- Alibaba Token Plan deepseek-v4.1-flashsubscription plan
- Alibaba Token Plan (China) deepseek-v4.1-flashMainland-China endpointsubscription plan
- DevPass (LLM Gateway) deepseek-v4.1-flashsubscription plan
- ClinePass cline-pass/deepseek-v4.1-flashsubscription plan
Data: models.dev (MIT), snapshot 2026-09-20. Prices are vendor-published; verify before purchase. Methodology.