October 21 update: Flash-1017 checkpoint
DeepSeek published an October 17 checkpoint of V4.1 Flash under the name DeepSeek-V4.1-Flash-1017. The official model card calls it a minor update, while reporting material gains in reasoning, coding, agent and visual tasks. The repository's initial-release history dates the checkpoint to October 17.
The vendor's table reports Terminal-Bench 4.0 rising from 31.2 to 42.3 and Humanity's Last Exam without tools from 36.8 to 42.6. These are DeepSeek's evaluations under the stated harnesses, not results independently reproduced by Signal. They do not establish a universal advantage on a customer's workload.
The model card licenses the repository and weights under MIT and supplies inference and prompt-format instructions. It describes reduced task execution time and output length, but realized cost depends on deployment and workload. The checkpoint remains part of the V4.1 Flash line rather than a new foundation-model generation.
The image retained for this canonical update is the exact benchmark graphic from the original September 10 announcement. It represents the same model family; its displayed scores are historical and are not the new Flash-1017 results. The September launch reporting below is preserved as dated background.
Original September launch coverage
DeepSeek published DeepSeek-V4.1 Flash on September 10, presenting it as the smallest model in a new architecture family. The company’s announcement says the model combines native multimodal vision with higher capability, faster inference, greater throughput, and a path to scaling into larger models. Reuters separately reported the launch and those design goals, providing independent confirmation that the release occurred.
An asymmetric architecture
DeepSeek describes V4.1 Flash as a 552-billion-parameter mixture-of-experts model built around a new Causal-Encoder-Decoder structure. Its announcement says only 8 billion input parameters and 16 billion output parameters are activated, a design it presents as a way to lower serving cost relative to models of similar total size. Those architecture and parameter figures come from DeepSeek and have not been independently audited here.
Cache and API changes
The company says the new model reduces KV-cache requirements to one-quarter of the previous model’s HBM demand and one-eighth of its SSD demand. DeepSeek also says V4.1 Flash is available through its API under the name deepseek-flash, with lower pricing taking effect on September 10; older V4 Flash routes are temporarily redirected, and the company plans to route deepseek-v4-pro requests to Flash after September 14 until a future V4.1 Pro release.
DeepSeek labels the announcement’s final section model open source and links a V4.1 Flash model page and technical report on Hugging Face. It says it will support inference adaptations for the open-source community and asks operators with large deployments to bring substantial GPU and storage resources. The release does not independently establish how broadly weights can be deployed, the realized cost reduction, or the reported benchmark advantage.