DeepSeek: DeepSeek V4 Flash Vision Exp
Provider: deepseek · deepseek/deepseek-v4-flash-vision-exp
How DeepSeek: DeepSeek V4 Flash Vision Exp pricing compares
At $0.66 / 1M output tokens, DeepSeek: DeepSeek V4 Flash Vision Exp is the 5th-cheapest of 14 deepseek models we track. It's cheaper than 74% of the 392 paid models on computetrail (102nd-cheapest overall). Output costs 3.0× its input price ($0.22 / 1M in). For a typical 3:1 input-to-output workload that works out to about $0.33 / 1M blended. Reading cached input costs 97% less than fresh input, so prompt caching cuts repeat-context cost sharply. Its 1,048,576 tokens context window ranks 32nd-largest of 392 models with a published limit.
Output price history
Across 6 daily snapshots since 2026-08-22, the output price has held steady at $0.66 / 1M — no change so far.
Similar-priced models
| Model | Provider | Output $/1M | Context |
|---|---|---|---|
| WizardLM-2 8x22B | microsoft | $0.62 | 65,535 |
| Meta: Llama 3.3 70B Instruct | meta-llama | $0.71 | 131,072 |
| Mistral: Saba | mistralai | $0.6 | 32,768 |
| Cohere: Command R (08-2024) | cohere | $0.6 | 128,000 |
| Mistral: Mistral Small 3.1 24B | mistralai | $0.555 | 128,000 |
| Qwen: Qwen-Plus | qwen | $0.78 | 1,000,000 |
Prices are recorded daily from public provider data via OpenRouter and normalized to USD per 1M tokens. See methodology.