DeepSeek V4 Flash Vision Exp
DeepSeek· released 21 Aug 2026
deepseek/deepseek-v4-flash-vision-exp- Input / 1M tokens
- $0.216
- Output / 1M tokens
- $0.647
Longest answer: 262K tokens
About
DeepSeek V4 Flash Vision Exp is an experimental vision-enabled version of DeepSeek V4 Flash 0731. It adds image understanding while matching the base model on text capabilities, though the listed summary cuts off before naming those capabilities. It accepts text and image inputs, returns text, and supports tools and reasoning. Its context window is 1,048,576 tokens, with a maximum output of 262,144 tokens. Input costs USD 0.2156 per 1M tokens and output costs USD 0.6468 per 1M tokens. The model is listed as open weights, with Hugging Face identifier deepseek-ai/DeepSeek-V4-Flash-Vision-Exp and model identifier deepseek/deepseek-v4-flash-vision-exp. For builders, the combination of image understanding, tools, and a very large context window may fit workflows that need visual and textual input together. The model was released on 2026-08-21.
Who it is for
Teams building text workflows that also need image understanding may consider this experimental model. Its large context window and tool support are relevant to agent-style applications.
What is good
- Accepts both text and image inputs
- 1,048,576-token context window
- Open weights listed
- Supports tools and reasoning
What to know first
- Experimental model
- Outputs text only
Verdict
DeepSeek V4 Flash Vision Exp adds image input to a text-capable model and supports tools and reasoning. Its experimental status and text-only output are important constraints to weigh.
Details
- Lab
- DeepSeekopenrouter.ai · 3 Oct 2026
- Context
- 1,049Kopenrouter.ai · 3 Oct 2026
- Input price
- $0.216 / 1Mopenrouter.ai · 3 Oct 2026
- Output price
- $0.647 / 1Mopenrouter.ai · 3 Oct 2026
- Max output
- 262,144 tokensopenrouter.ai · 3 Oct 2026
- Inputs
- text, imageopenrouter.ai · 3 Oct 2026
- Open weights
- Yesopenrouter.ai · 3 Oct 2026
More newest models
See the list
Listed on Inferse
- Newest Models in 2026466 listed
- Longest-Context Models in 2026466 listed
- Cheapest Language Models in 2026437 listed
- Models with Tool Calling in 2026395 listed
- Reasoning Models in 2026332 listed
- Vision Language Models in 2026295 listed
- Open-Weight Models in 2026178 listed
- DeepSeek Models: Pricing per 1M Tokens and Context (2026)18 listed
Sources
- openrouter.ai/deepseek/deepseek-v4-flash-vision-exp· checked 3 Oct 2026