deepseek-v4.1-flash-1017
DeepSeek-V4.1-Flash-1017 is the latest iteration of V4.1-Flash, trained with self-feedback-based reinforcement learning to deliver substantial gains over its predecessor across benchmarks such as Terminal-Bench 4.0, DeepSWE v1.1, and ProgramBench, while shortening task runtime and output length to reach a new performance–cost Pareto optimum across multiple harnesses. Its ability to build complete projects from scratch is further strengthened, enabling "wish-based" programming tasks such as game
Context
1M
Input / 1M
$0.30
Output / 1M
$1.20
Modalities
Protocol
3- chat.completions
- messages
- responses
Status & performance
Last 7 days
Loading model performance
Capabilities
TextVisionLong contextCache
Pricing
Time-based pricing
Weekly schedule · Asia/ShanghaiWhen matched, the complete price for that range replaces the default; missing items are not filled from it.
00:0006:0012:0018:0024:00
Default pricing
Applies when no time range matches on Every dayInput tokens
$0.3/1M
Output tokens
$1.20/1M
Cache read
$0.03/1M
Cached input
$0.03/1M
Prices in USD per 1M tokens unless noted otherwise. Batch calls and cache hits receive additional discounts; live rates apply in the Workspace.