DeepSeek Releases DeepSeek-V4.1-Flash-1017 Under MIT License
The coding benchmark chart and the MIT license put the price question on the table, since DeepSeek is selling usage cheaply while giving the weights away.
Reporting from 1 source: GIGAZINE.
DeepSeek released the open model DeepSeek-V4.1-Flash-1017 on October 17, 2026, positioned as an updated, faster version of DeepSeek-V4.1-Flash. On the DeepSWE v1.1 coding agent benchmark it scores above GPT-6 Astra, GPT-6.1 Sol and Claude Fable 5 while keeping cost low, and it recorded the highest scores on multiple tests. Peak pricing is $0.3 per million input tokens and $1.2 per million output tokens, halved off-peak.
The model has 763 billion parameters, which DeepSeek says makes it impractical on a personal computer. Running it at usable speeds therefore depends on the official API or a third-party API, and Ollama and LM Studio have begun hosting it on servers in the US.
Peak-hour rates apply Monday through Friday, 1:00 to 4:00 and 6:00 to 10:00 UTC: $0.3 input, $0.006 cached input and $1.2 output per million tokens, with off-peak priced at half. The weights are downloadable from Hugging Face and ModelScope under the MIT License.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.