DeepSeek V4.1 Flash
Provider: DeepSeek CN
Input price
$0.21
/ 1M tokens
Output price
$0.63
/ 1M tokens
Cache price
$0.0063
/ 1M tokens (read)
Batch API
$0.11 / $0.34
input / output per 1M tokens
Context
1.05M tokens
Max output 393K tokens
Overview
DeepSeek V4.1 Flash is a budget reasoning and vision model built on a causal encoder-decoder architecture. It targets high-context multimodal tasks requiring structured outputs.
Strengths
- Activates 8B parameters on input and 16B on output
- Supports 1,048,576 token context window and 943,718 max output
- Includes vision, function-calling, json-mode, and thinking capabilities
- Offers cached input at $0.0063 per million tokens
Best use cases
- Processing large multimodal documents with vision
- Executing structured JSON generation tasks
- Running complex multi-step reasoning workflows
Pricing position
Input is priced at $0.21 and output at $0.63 per million tokens, matching Qwen3 VL 235B on input while delivering lower output costs than MiniMax M2.7.
Caveats
Has no free tier available.
Notes drafted by an LLM from the figures on this page (2026-11-13) and rewritten when the price changes.
Key facts
- Price basis
- Median third-party host price (no first-party API)
- Max output
- 393K tokens
- Listed on OpenRouter
- 2026-09-10
- Tracked since
- 2026-09-10
- Last updated
- 2026-11-14
Features
Image inputFunction callingJSON outputReasoning
Price history
- 2026-11-13 Price change $0.039 / $1.20 → $0.21 / $0.63
- 2026-11-12 Price change $0.20 / $0.78 → $0.039 / $1.20
- 2026-11-12 Price change $0.21 / $0.63 → $0.20 / $0.78
- 2026-11-10 Price change $0.20 / $0.78 → $0.21 / $0.63
- 2026-09-10 Added
Monthly cost estimate
| Scale | Input (M) | Output (M) | Monthly |
|---|---|---|---|
| 0.1M (hobby) | 0.1 | 0.05 | $0.05 |
| 1M (small app) | 1 | 0.5 | $0.53 |
| 10M (mid scale) | 10 | 5 | $5.25 |
| 100M (large) | 100 | 50 | $52.50 |
※ Cache and batch discounts not factored in. Detailed calculator
Official resources
Other models from DeepSeek
Similar-priced models
Source: OpenRouter API (deepseek/deepseek-v4.1-flash)