DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis | Hacker News

Hacker News commenters debated DeepSeek V4 Flash 0731’s benchmarks, response quality, and inference speed.

Some praised its strong performance at a far lower cost than competing models like OpenAI’s offerings.

Others pointed to limits in coding and agent use, plus concerns about censorship, safety, and missing multimodal support.

The thread underscored how fast-improving open-weight models are reshaping expectations for price, speed, and model choice.