OpenAI reveals Sora, a text-to-video model producing startlingly realistic 1-minute clips. The generative video moment arrives, and every studio rethinks its stack.

Anthropic releases Claude 3 in three tiers: Opus, Sonnet, and Haiku. Opus claims GPT-4-class performance; Haiku delivers near-instant responses at low cost.

Meta releases Llama 3 in 8B and 70B sizes. Open-source LLMs reach performance parity with mid-tier closed models. The local-AI movement accelerates.