AnthropicAI has announced the release of Claude 3 Haiku, positioning it as the fastest and most affordable model in its intelligence class, now generally available for Claude Pro users. Claude 3 Haiku is distinguished by its speed, processing up to 21,000 tokens (approximately 30 pages) per second for prompts under 32,000 tokens, making it three times faster than its competitors for the majority of workloads. Additionally, its affordability is highlighted by offering 4M prompt tokens for just $1, enabling widespread integration across various applications. Claude 3 Haiku also introduces image-to-text vision capabilities and supports multiple languages, expanding its use cases beyond those feasible with previous models due to constraints on price, latency, and functionality. The model is available on platforms such as Poe, Perplexity Labs (for free), and Amazon Bedrock, boasting increased steerability in a 200k context window. This release is seen as a significant advancement in the AI field, promising to unlock many new applications and enhance the speed and cost-effectiveness of AI solutions.