OpenAI and Anthropic cut their LLM prices by up to 80% within a few months. Why?
Points and comments are a snapshot, not live.
OpenAI and Anthropic slashed LLM prices up to 80% due to open-weight competition and commoditization.
From June to September 2026, OpenAI and Anthropic cut LLM prices by up to 80% on models like GPT-5.6 Luna and GPT-6.1 Sol, driven by competition from open-weight models such as DeepSeek V4 Flash and Kimi K3.1 Pro. Cache read cost reductions (75-99% hit rate) further lowered effective prices. The author argues that token prices reflect genuine competition among dozens of inference providers, not loss-leading; prices are expected to continue falling toward marginal cost. Both companies are nearing IPOs, and subscription plans have had usage allowances cut to offset cheaper per-token costs.
What commenters are saying
Commenters largely agree that competition is driving prices down, with some noting parallels to mobile data commoditization. Several share practical tips: Claude Code and Zed Agent work well with Chinese models like GLM 5.3 and DeepSeek 4.1 after hitting refusal blocks on US models. A minority question financial sustainability, arguing frontier labs are burning cash and may implode or require bailouts. Others counter that Amazon and Apple precedents show prolonged losses can precede profitability.