Home / DeepSeek / DeepSeek V4 Flash (0731 snapshot)
DeepSeek V4 Flash (0731 snapshot) API price
deepseek-v4-flash-0731
The cheapest DeepSeek V4 Flash (0731 snapshot) API is $0.056/M input at InferX (inference host) - all 7 providers compared below, per 1M tokens.
Pinned snapshot of DeepSeek V4 Flash - this build does not change; the rolling name moves on without it.
SmallDeepSeek1.0M context7 sellers
Input
$0.056
per 1M tokens · InferX (inference host)
up to $0.44
Output
$0.112
per 1M tokens · InferX (inference host)
up to $1.32
Where to buy it
Cheapest DeepSeek V4 Flash (0731 snapshot) API providers compared - 7 sellers, from $0.056/M.
| Provider | Input $/M | Output $/M | Cached in | TPS |
|---|---|---|---|---|
| InferX (inference host)60% off | 0.056cheapest | 0.112 | 0.011 | - |
| Nahcrof (inference host)Q8_0 | 0.080 | 0.100 | 0.0030 | 86 |
| GMI Cloud (featured list)20% off | 0.112 | 0.224 | - | - |
| RunInfra (inference host) | 0.130 | 0.270 | 0.010 | 260 |
| Together AI (inference host) | 0.140 | 0.280 | 0.030 | - |
| OrcaRouter (aggregator) | 0.147 | 0.295 | - | - |
| TokenRouter (aggregator) | 0.440 | 1.32 | 0.014 | - |
7.9× between the cheapest and the dearest of 7 sellers.
Against its tier
- This model$0.168
- Cheapest · Qwen 3.7 Flash$0.16
- Tier median · GPT 5.6 Luna Pro$0.70
- Dearest · Claude Haiku 4.5$6.00
#2 of 17 small · cheaper than 94%
Details
- Context window
- 1.0M1,048,576 tokens
- Developer
- DeepSeek
- Intelligence
- 51.8Reasoning, Max Effort
- Sellers
- 77.9× spread
- Cached input
- $0.0112where the source publishes one
- Cheapest at
- InferX (inference host)
- Combined
- $0.168in + out per 1M
- Last checked
- 2026-08-28
- Throughput
- 86 tok/sas published by Nahcrof (inference host)