On July 30, 2026, OpenAI slashed prices for its lower-tier GPT-5.6 Luna and Terra models by 80% and 20% respectively, while keeping its flagship Sol model price unchanged. The move comes amid rising enterprise scrutiny over ballooning AI costs and intense competition from cheaper Chinese rivals like Z.ai's GLM-5.2. The price cuts also pressure Anthropic, whose Claude Sonnet 4.6 remains more expensive than OpenAI's mid-tier Terra model.
OpenAI lower-tier model discounts
- ▪OpenAI attributed the price reductions to efficiency gains from GPT-5.6, including code improvement and performance optimization during internal development
- ▪On July 30, 2026, OpenAI reduced the price of its smaller GPT-5.6 Luna model by 80% and its mid-tier Terra model by 20%
- ▪OpenAI left the pricing of its flagship Sol model unchanged during the price cuts announced on July 30, 2026
Chinese AI rivals pressure
- ▪OpenAI and Anthropic face competition from open-source Chinese rivals, such as Z.ai's GLM-5.2 model, which offer similar performance at lower costs
- ▪The price cuts by OpenAI on July 30, 2026, come as United States artificial intelligence companies compete against cheaper Chinese rivals
Enterprise cost scrutiny rising
- ▪Businesses are increasingly scrutinizing ballooning artificial intelligence costs and facing hefty, unpredictable bills due to a shift toward usage-based pricing
- ▪Analysts suggest that while cutting prices may boost usage, it could strain the finances of artificial intelligence companies ahead of anticipated initial public offerings
Anthropic competitive positioning
- ▪OpenAI's price cuts increase competitive pressure on Anthropic, whose Claude models dominate enterprise and developer use but are positioned at the costlier end of the market
- ▪Anthropic's mid-tier Claude Sonnet 4.6 model costs $3 per million input tokens and $15 per million output tokens, which is higher than OpenAI's newly discounted Terra model rates
Token pricing structure changes
- ▪Under OpenAI's new pricing, sending text to the Luna model drops to $0.20 per million tokens and generating responses falls to $1.20
- ▪Overall token prices have decreased over the past year, but the total cost of completing tasks is rising as artificial intelligence firms shift from flat subscriptions to usage-based pricing
Story comments
Loading comments…