DeepSeek Opens Limited-Time Beta of V4.1Flash: New Architecture, Native Multimodal, Expires Sept 10
APPSO · wechat · 2026-09-08
DeepSeek has opened a limited-time beta of DeepSeek-V4.1Flash, an intermediate test version with a new model architecture and native multimodal support. Developers can call it via the existing baseurl with model name deepseek-v4.1-flash-expires-on-0910; pricing matches V4Flash, capped at 20 concurrent requests per account, and the model auto-expires on September 10.
DeepSeek's feedback survey explicitly asks whether it can "fully replace V4Pro," signaling the model aims to bring Pro-level capability at Flash speed and cost. APPSO's hands-on tests found strong Chinese long-form writing, a polished polaroid-camera web demo built in 5 minutes, a middling macOS-style web OS, and a spaceship HUD demo showcasing refined per-layer UI responsiveness. Generation speed is the standout, especially valuable for iterative coding workflows.
More from Models
- Kimi K3.1 Pro tops MathArena open models, closing frontier gap to about 6 months — teortaxesTex · 2026-10-26
- Ruff creator Charlie Marsh: frontier models find infinite bugs in any software — charliermarsh · 2026-10-26
- Dev testing one skill across Opus 5.5 and Sol 6.1 finds instruction following 'all over the map' — lucasmeijer · 2026-10-26
- supermemory unveils learner-1, claiming continual learning for any model — julianweisser · 2026-10-26
- Anthropic says Opus 5.5 cuts cost per task by ~40% vs Opus 5 — taufiqintech · 2026-10-26
- AI search APIs ranked: top 12 within 7 points, yet prices differ nearly 7x — FinanceYF5 · 2026-10-26