AI Costs Soar: Price Hikes Rock Industry

The AI industry is facing a sharp increase in operational costs, leading to higher prices for consumers. Anthropic's new Fable 5 model, released earlier this week, now costs twice as much as its predecessor, Opus 4.8, despite some users reporting issues with its safety guardrails. This price hike comes as competition intensifies and infrastructure demands grow.

In response to these rising costs and user concerns, OpenAI is reportedly considering reducing the prices it charges for tokens, the basic units of AI usage. This move could alleviate some of the financial pressure on customers. Meanwhile, Google's Gemini 3.5 Flash, launched last month, presents itself as the most budget-friendly option, with significantly lower API costs per token.

Fable 5's pricing shifts to a pay-as-you-go model starting June 23, charging $10 per million input tokens and $50 per million output tokens. This model significantly impacts users engaging in complex, long-running tasks like coding. GPT-5.5 Pro, available via API, costs $5 per million input tokens and $30 per million output tokens, offering a batch tokenization discount. Gemini 3.5 Flash remains accessible at $1.50 per million input and $9 per million output tokens, positioning it as a highly competitive choice for developers and users seeking cost-effectiveness.