Model pricing
DeepSeek: DeepSeek V3.1 API Pricing
Compare 9 providers offering DeepSeek: DeepSeek V3.1. Cheapest quote is $0.250 per 1M input tokens.
- Providers
- 9
- Official
- —
- Cheapest
- $0.250
- Context
- 164K
Provider pricing
Cheapest and most expensive quotes differ by 2.6×.
Estimate my cost
Period
based on 30 days
| Provider | Input | Output | Cached input | Context | vs official | Your cost | |
|---|---|---|---|---|---|---|---|
17.3M | $0.250 | $0.950 | $0.130 | 164K | — | $13cheapest | Use |
DeepInfraInference platform | $0.250 | $0.950 | $0.130 | 164K | — | $13cheapest | |
NovitaInference platform | $0.270 | $1.00 | $0.135 | 131K | — | $14+$0.900 | |
SiliconFlowInference platform | $0.270 | $1.00 | — | 164K | — | $14+$0.900 | |
AtlasCloudInference platform | $0.300 | $0.950 | $0.130 | 131K | — | $15+$1.50 | |
CoreWeaveInference platform | $0.550 | $1.65 | $0.550 | 161K | — | $26+$13 | |
GoogleInference platform | $0.600 | $1.70 | — | 164K | — | $28+$15 | |
MaraInference platform | $0.600 | $1.70 | — | 131K | — | $28+$15 | |
SambaNovaInference platform | $0.650 | $1.50 | — | 131K | — | $29+$15 |
Prices captured Jul 26, 2026
Prices change frequently and promotional rates expire — always confirm on the provider's own pricing page before committing. Rates far below official pricing often indicate shared or resold accounts.