Model Library
DeepSeekReleased April 24, 2026DeepSeek V4 Flash (max)
DeepSeek · Released April 24, 2026
Intelligence score
40
At a glance
- Intelligence
- 40
- Speed
- 113.0tokens/sec
- Latency
- 1.26stime to first token
- Input price
- $0.14per 1M tokens
- Output price
- $0.28per 1M tokens
- Context window
- 4tokens
- Parameters
- 284billions
Summary
Comparison summary
DeepSeek V4 Flash (Reasoning, Max Effort) scores 40 on the Artificial Analysis Intelligence Index, placing it well above average among other open weight models of similar size (median: 25). DeepSeek V4 Flash (Reasoning, Max Effort) generates output at 113.0 tokens per second (based on DeepSeek's API), which is well above average compared to other open weight models of similar size (median: 64.8 t/s). DeepSeek V4 Flash (Reasoning, Max Effort) costs $0.14 per 1M input tokens (very competitive, median: $0.59) and $0.28 per 1M output tokens (very competitive, median: $2.20), based on DeepSeek's API.
Specifications
Technical specifications
| Model name | DeepSeek V4 Flash (max) |
|---|---|
| Provider | DeepSeek |
| Release date | April 24, 2026 |
| Intelligence score | 40 |
| Speed | 113.0 tokens/sec |
| Latency (TTFT) | 1.26s |
| Input price | $0.14 / 1M tokens |
| Output price | $0.28 / 1M tokens |
| Context window | 4 tokens |
| Parameters | 284B |
| Reasoning | Yes |
| Input modalities | text |
| Output modalities | — |
| Open weights | Yes |
| License | MIT |
| API availability | — |
| Source | Model Intelligence legacy profile: deepseek-v4-flash |
Pricing
Token prices
Price per 1M tokens
DeepSeek V4 Flash (max), as listed in the catalog record
- Input$0.14
- Output$0.28
| Token type | Price per 1M |
|---|---|
| Input | $0.14 |
| Output | $0.28 |
Model Intelligence legacy profile: deepseek-v4-flash
FAQ · 21
Frequently asked questions
DeepSeek V4 Flash (Reasoning, Max Effort) was released on April 24, 2026.
DeepSeek V4 Flash (Reasoning, Max Effort) was created by DeepSeek.
DeepSeek V4 Flash (Reasoning, Max Effort) scores 40 on the Artificial Analysis Intelligence Index, placing it well above average among other open weight models of similar size (median: 25).
DeepSeek V4 Flash (Reasoning, Max Effort) generates output at 113.0 tokens per second (based on DeepSeek's API), which is well above average compared to other open weight models of similar size (median: 64.8 t/s).
DeepSeek V4 Flash (Reasoning, Max Effort) has a time to first token (TTFT) of 1.26s (based on DeepSeek's API), which is very competitive compared to other open weight models of similar size (median: 2.03s).
DeepSeek V4 Flash (Reasoning, Max Effort) costs $0.14 per 1M input tokens (very competitive, median: $0.59) and $0.28 per 1M output tokens (very competitive, median: $2.20), based on DeepSeek's API.
DeepSeek V4 Flash (Reasoning, Max Effort) costs $0.14 per 1M input tokens and $0.28 per 1M output tokens (based on DeepSeek's API). For a blended rate (7:2:1 cache hit/input/output ratio), this is $0.06 per 1M tokens. Pricing may vary by provider.
When evaluated on the Intelligence Index, DeepSeek V4 Flash (Reasoning, Max Effort) generated 230M output tokens, which is at the higher end compared to other open weight models of similar size (median: 92M).
Yes, DeepSeek V4 Flash (Reasoning, Max Effort) is a reasoning model. It uses extended thinking or chain-of-thought reasoning to work through complex problems before providing an answer.
DeepSeek V4 Flash (Reasoning, Max Effort) supports text input.
DeepSeek V4 Flash (Reasoning, Max Effort) supports text output.
No, DeepSeek V4 Flash (Reasoning, Max Effort) does not support image input. It can only process text.
No, DeepSeek V4 Flash (Reasoning, Max Effort) is not multimodal. It only supports text input.
DeepSeek V4 Flash (Reasoning, Max Effort) has a context window of 1.0M tokens. This determines how much text and conversation history the model can process in a single request.
Yes, DeepSeek V4 Flash (Reasoning, Max Effort) is open weights. The model weights are publicly available and can be downloaded for self-hosting.
DeepSeek V4 Flash (Reasoning, Max Effort) has 284 billion parameters (13 billion active).
DeepSeek V4 Flash (Reasoning, Max Effort) is a Mixture of Experts (MoE) model with 284 billion total parameters, but only 13 billion active parameters are used during inference.
DeepSeek V4 Flash (Reasoning, Max Effort) is released under the MIT license. This license allows commercial use.
DeepSeek V4 Flash (Reasoning, Max Effort) achieves a score of 40 on the Artificial Analysis Intelligence Index. This composite benchmark evaluates models across reasoning, knowledge, mathematics, and coding.
Yes, DeepSeek V4 Flash (Reasoning, Max Effort) is available via API through 7 providers.
DeepSeek V4 Flash (Reasoning, Max Effort) is available through 7 API providers.