Tag: #llm-pricing12 results
This week's roundup covers DeepSeek's custom inference chips, OpenAI's IPO rumors, and significant price competition from Chinese AI models. It also highlights new model developments from MiniMax and xAI.
Volcengine has launched the Doubao 2.1 model series, featuring significant price reductions for the Pro and Turbo versions. The company also clarified it has no current plans to spin off or list the cloud division independently.
Zhipu AI open-sourced its flagship GLM-5.1 model while raising API prices by 10%. This marks the second price increase this year, after a February overhaul that hiked coding subscription rates over 30%. The moves signal a push to monetize AI amid intensifying US rivalry.
Opus 5 has demonstrated superior performance compared to Fable 5 while maintaining a significantly lower cost structure. This shift challenges the existing pricing strategy of Anthropic in the competitive LLM market.
Moonshot AI has introduced the Kimi K3 model, priced significantly higher than competitors like DeepSeek. This shift indicates a move toward high-value, complex task capabilities.
Alibaba has introduced aggressive price cuts of up to 80% for its Qwen3.7-Max and Qwen3.7-Plus models on the Qoder platform. This strategic move targets global developers and aims to compete directly with US-based AI providers like Anthropic.
The era of aggressive AI token subsidies is ending as structural imbalances between startups and tech giants become apparent. Google's potential 80% price cut is forcing a market-wide pricing correction.
Step-1 Flash model significantly improves efficiency, with task costs reduced to just 1/9th of Claude 3 Opus. The update positions the model as a high-performance solution rather than just a budget alternative.
DeepSeek has officially made its 75% discount on the V4 Pro model permanent, moving past the original May 31 deadline. This aggressive pricing strategy significantly undercuts major competitors, signaling a deepening price war in the LLM market.
Anthropic has extended free access to its Claude Fable model until July 19 for paid subscribers, while signaling a future pay-per-use model. Analysts suggest this strategy aims to capture user data and market share amidst aggressive competition from OpenAI.