The company claimed that it had greatly reduced inference costs for AI agents by reducing this memory footprint to 890 bytes ...

Developer’s V4.1 Flash launch highlights China’s AI battle, where efficiency and speed drive competition amid chip curbs.

DeepSeek V4.1-Flash has 552B total parameters but activates 8B during prefill and 16B during decode. Here's why the ...

DeepSeek says the new model improves coding and agentic workloads while giving users control over its reasoning effort.

HONG KONG/SINGAPORE, Sept 9 (Reuters) - Chinese artificial intelligence startup DeepSeek has tapped CITIC Securities to ...

Flash, its smallest model in a new architecture family, as the Chinese AI startup prepares for a Shanghai STAR Market IPO.

DeepSeek-V4.1-Flash enters testing phase, potentially impacting rankings. Anthropic best AI model by September 2026 at 87.5% YES.

Delivering higher efficiency and reduced KV cache consumption, this new open-source model outperforms several flagship ...

DeepSeek's V4.1 Flash model packs 552 billion parameters at a fraction of rivals' costs, sending competitor stocks tumbling ...

Chinese artificial intelligence lab DeepSeek has officially rolled out its V4.1-Flash model, marking a significant milestone ...

DeepSeek’s V4.1 Flash AI model offers groundbreaking affordability, challenging Anthropic, OpenAI, and Chinese competitors in ...

DeepSeek V4.1 Flash processes up to 400 tokens per second in this temporary test build. See how the fast, affordable model ...