According to @godofprompt, xAI has released Grok-4-Fast, a next-generation AI model featuring a 2 million-token context window and advanced efficiency engineering. The model offers built-in reasoning, function calling, structured outputs, and a dedicated non-reasoning mode for high-throughput scenarios. Pricing is highly competitive at $0.20 per million tokens for input and $0.50 for output, making it six times cheaper than Grok-4-0709 with similar performance. Grok-4-Fast also introduces prompt caching, allowing users to pay once and reuse prompts indefinitely. Tools such as Web Search, X Search, Code Execution, and Doc Search are free until November 21, 2025, after which they will cost $10 per 1,000 calls. This release marks a significant step in commoditizing advanced AI reasoning, providing businesses with affordable, long-context, low-latency AI solutions (Source: @godofprompt on Twitter).

According to Andrej Karpathy on Twitter, many users remain unaware that ChatGPT's o3 model is currently the superior option for complex reasoning and professional applications compared to the newer 4o model. Karpathy emphasizes that o3 delivers significantly better performance on important or difficult tasks, making it the preferred choice for enterprise and advanced use cases where accuracy and logical reasoning are critical (source: @karpathy, June 2, 2025). Businesses leveraging ChatGPT for professional workflows should prioritize o3 to maximize outcomes and reliability.