AMD Radeon Cloud DeepSeek V4 Flash Vision
Official AMD endpoint for DeepSeek V4 Flash Vision (Exp) multimodal reasoning with 1M context.
Sobre o modelo
Why Choose DeepSeek V4 Flash Vision (Exp)?
DeepSeek V4 Flash Vision (Exp) is DeepSeek's experimental multimodal foundation model running on AMD Radeon Cloud GPU clusters, combining sub-second latency with 1M context and high-fidelity visual reasoning.
Key Strengths & Capabilities
- 1M Context Window: Ingest massive multi-image documents, UI mockups, and extensive multimodal prompts without truncation.
- High-Throughput Vision Reasoning: Accelerated by AMD Instinct/Radeon GPU architecture with dynamic sglang-router dispatching.
- Dual Protocol Support: Seamlessly switch between standard OpenAI Chat Completions and Anthropic Messages formats.
- Free Developer Access: Freely accessible in AMD Radeon Cloud Token Factory without requiring credit card registration.
Free Tier Basis
Offered as a Public Free Model API in AMD Token Factory. Points indicate relative usage rather than billing charges.
Como Acessar Gratuitamente (via AMD Radeon Cloud)
Why Choose AMD DeepSeek V4 Flash Vision?
AMD GPU Cloud served multimodal vision model supporting 1M tokens context for comprehensive image, diagram, and code reasoning.
Experimente no navegador
Escolha um modelo, cole sua própria chave gratuita e execute. Sua chave é enviada uma vez para chamar o provedor e nunca é armazenada em nossos servidores.
Escolha um modelo, cole sua própria chave gratuita e execute. Sua chave é enviada uma vez para chamar o provedor e nunca é armazenada em nossos servidores.
Exemplos de Código
curl https://developer.amd.com.cn/radeon/api/v1/chat/completions \
-H "Authorization: Bearer $AMD_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "DeepSeek-V4-Flash-Vision-Exp",
"messages": [
{
"role": "user",
"content": [
{"type": "text", "text": "Describe what is in this image:"},
{"type": "image_url", "image_url": {"url": "https://example.com/demo.png"}}
]
}
]
}'
Feedback da comunidade
Os comentários estão vinculados à sua conta GitHub — entre para participar da discussão.