DeepSeek · DeepSeek family
deepseek-v4-flash-vision-exp
Ask it anything, hand it a long document, or have it write and edit for you.
Reads imagesTool callingThinks before it answersLong documents
Context window
1Mtokenshow much it can read at onceMax output
393Ktokenslongest answer in one goInput price
$0.15/ 1Mper million tokens you sendOutput price
$0.60/ 1Mper million tokens it writesCached input
$0.003/ 1Mper million tokens it has seen beforePrices and specs come from the provider.
In plain words
What it is, and when to use it
deepseek-v4-flash-vision-exp is made by DeepSeek, and is part of its DeepSeek family. Ask it anything, hand it a long document, or have it write and edit for you. It can read about 1,573 pages in one go.
Try these
Read my rental agreement and tell me which parts could work against me.Try itHelp me write a birthday message for my mom — warm and sincere, not cheesy.Try itI'm thinking about changing jobs. Lay out the pros and cons of each choice for me.Try it
Two words you’ll see: a token is how AI counts text — 1,000 tokens is roughly 750 English words. The context window is how much it can read in one go.
For builders
Spec sheet
| Model ID | deepseek-v4-flash-vision-exp |
|---|---|
| Provider | DeepSeek |
| Family | DeepSeek |
| Input | Text, Images |
| Output | Text |
| Context window | 1,048,576 tokens |
| Max output | 393,216 tokens |
| Input price | $0.15 / 1M |
| Output price | $0.60 / 1M |
| Cached input | $0.003 / 1M |
Capabilities
- Reads images
- Tool / function calling
- Structured output (JSON)
- Extended thinking
- Prompt caching
- Makes images
- Audio in or out
Call it
curl https://api.wayjet.ai/v1/chat/completions \
-H "Authorization: Bearer $WAYJET_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "deepseek-v4-flash-vision-exp",
"messages": [{"role": "user", "content": "Hello"}]
}'Compare
Side by side
| Model | Context | Input / 1M | Output / 1M | Try it |
|---|---|---|---|---|
| deepseek-v4-flash-vision-exp | 1M | $0.15 | $0.60 | Try it |
| gpt-4o-mini | 128K | $0.15 | $0.60 | Try it |
| MiniMax-M3.1-Flash | 1M | $0.15 | $0.60 | Try it |
| mistral.ministral-3-8b-instruct | 128K | $0.15 | $0.15 | Try it |
Questions
About deepseek-v4-flash-vision-exp
Is deepseek-v4-flash-vision-exp free to try?+
Yes. Like every model on WayJet, it can be tried free once you sign up. After that you pay as you go, at the provider’s price.
How much does deepseek-v4-flash-vision-exp cost?+
It is charged per token at the provider’s own price — see Input price and Output price above. WayJet adds one platform fee when you top up, with no monthly fee.
How much can deepseek-v4-flash-vision-exp read at once?+
See Context window above. As a rough guide, 1,000 tokens is about 750 English words, or one and a half pages.
Can I use deepseek-v4-flash-vision-exp in Claude Code or Codex?+
Yes. Create a WayJet API key and use the model ID deepseek-v4-flash-vision-exp. The API reference shows how to set up both.
Which is better, deepseek-v4-flash-vision-exp or gpt-4o-mini?+
It depends on the job. The table above sets their prices side by side; the quickest way to know is to ask both the same question in the Playground.
