DeepSeek Launches V4.1-Flash: 552B Multimodal Model with Cheaper API

DeepSeek released V4.1-Flash, a 552B-parameter multimodal model with native vision understanding that topped Terminal-Bench 2.1 at 90.6, alongside API price cuts. Third-party gateway Merge has already integrated the model.

2026-09-12 ~ 2026-09-14 · 3 related posts

• DeepSeek V4.1 Flash lands on Merge Gateway, tops Terminal-Bench 2.1 at 90.6 — shensi · 2026-09-12
• DeepSeek launches V4.1-Flash: 552B params, multimodal, faster, with API price cuts — emmanuelvivier · 2026-09-13
• DeepSeek releases V4.1-Flash: native vision, smallest of new family, credits automated RL environments — SergioPaniego · 2026-09-14