MiniMax and DeepSeek Launch Faster, Cheaper AI Models, Intensifying Competition

Key Intelligence

• ■MiniMax unveiled its new M3.1 Flash AI model, significantly reducing API prices and AI agent costs by up to 70% while improving performance.
• ■DeepSeek introduced new AI models in its Flash series and an open-source inference tool aimed at cutting AI operating costs and facilitating scalable model deployment.
• ■Both companies are directly challenging flagship-tier generative AI models from industry giants like OpenAI and Anthropic.
• ■These releases highlight a growing trend toward more accessible, cost-effective, and faster AI models, putting pressure on established LLM providers.
• ■The advancements suggest a diminishing "moat" for leading AI companies as smaller and open-source models become increasingly capable and prevalent.