alibaba:deepseek/deepseek-v4-flash
Common Name: DeepSeek V4 Flash (Alibaba Cloud)
DeepSeek V4 Flash hybrid-thinking model hosted on Alibaba Cloud Model Studio.
Specifications
Context
1000K
Maximum Output
384K
Inputtext
Outputtext
Performance (7-day Average)
Collecting…
Collecting…
Collecting…
Pricing
Input$0.1104/MTokens
Cached Input$0.0224/MTokens
Cache Creation$0.138/MTokens
Output$0.22/MTokens
Performance Metrics (24h)
Similar Models
$0.1104/$0.22/M
ctx1.0Mmax384Kavail—tps—
InOutCap
DeepSeek V4 Flash 0731 weight snapshot hosted on Alibaba Cloud Model Studio.
$1.485/$4.4559/M
ctx1.0Mmax131Kavail—tps—
InOutCap
Alibaba's Qwen3.8 flagship with hybrid thinking (default on) and vision support. 1M context, 131K max output.
$0.99/$1.9806/M
ctx1.0Mmax384Kavail—tps—
InOutCap
DeepSeek V4 Pro hybrid-thinking model hosted on Alibaba Cloud Model Studio.
$0.636/$1.908/M
ctx1.0Mmax384Kavail—tps—
InOutCap
DeepSeek V4 Pro 0813 (official weights) hosted on Alibaba Cloud Model Studio.