DeepSeek V4 Flash Vision Exp
by AMD · family deepseek-flash · listed Aug 2026RTVW
Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work
Current rates USD per 1M tokens
Inputfree
Outputfree
Cache readfree
Cache write—
Against the market
On the record
- Model id
DeepSeek-V4-Flash-Vision-Exp- Context window
- 1M tokens
- Max output
- 384K tokens
- Input modalities
- text, image
- Output modalities
- text
- Reasoning
- yes
- Tool calls
- yes
- Open weights
- yes
- Knowledge cutoff
- —
- Released
- 2026-08-21
- Last updated
- 2026-09-01
- Provider docs
- developer.amd.com.cn/radeon/tokenfactory
Back of the envelope
≈
Also listed at
- OpenCode$0.14 in · $0.28 out
- Eden AI$0.15 in · $0.60 out
- CommandCode$0.15 in · $0.60 out
- DeepSeek$0.15 in · $0.60 out
- AIHubMix$0.155 in · $0.62 out
- Vercel AI Gateway$0.216 in · $0.647 out
- NanoGPT$0.216 in · $0.647 out
- OrcaRouter$0.22 in · $0.66 out
- CrossModel$0.30 in · $1.20 out
- OpenRouter$0.44 in · $1.32 out
- Hugging Face$0.44 in · $1.32 out
- AI/ML API$0.44 in · $1.32 out
- Kilo Gateway$0.44 in · $1.32 out
- SiliconFlow$0.44 in · $1.32 out
- Deep Infra$0.44 in · $1.32 out