Users are excited about DeepSeek's 90% cached input price cut because it boosts cost-effectiveness and usefulness for agentic AI applications, prompting consideration of Grok 4.5 as a default model.
Based on 1 visible X reactions from 3 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@ns123abc I'm tempted to make Grok 4.5 the default backing model for agents on http://Autohive.com with my experience this weekend. I've already converted my personal agents over and quality has remained high, while costs are down. Great model.
When DeepSeek cut cached input prices by 1/10th they made V4 Pro 10x more relevant for agentic workflows If Grok 4.5 violently cuts cached prices, it becomes the default most used* agentic model overnight https://x.com/ns123abc/status/2076440896661266751/photo/1 https://twitter.com/elonmusk/status/2076411423563948142
@elonmusk @Alezander907 DeepSeek proved that slashing cached prices 10x makes their model a lot more useful (despite lagging behind frontier). If Grok does the same, it becomes the most used model for agent meta by default, is my prediction
@ns123abc I'm tempted to make Grok 4.5 the default backing model for agents on http://Autohive.com with my experience this weekend. I've already converted my personal agents over and quality has remained high, while costs are down. Great model.
When DeepSeek cut cached input prices by 1/10th they made V4 Pro 10x more relevant for agentic workflows If Grok 4.5 violently cuts cached prices, it becomes the default most used* agentic model overnight https://x.com/ns123abc/status/2076440896661266751/photo/1 https://twitter.com/elonmusk/status/2076411423563948142
@elonmusk @Alezander907 DeepSeek proved that slashing cached prices 10x makes their model a lot more useful (despite lagging behind frontier). If Grok does the same, it becomes the most used model for agent meta by default, is my prediction
Users are excited about DeepSeek's 90% cached input price cut because it boosts cost-effectiveness and usefulness for agentic AI applications, prompting consideration of Grok 4.5 as a default model.
Based on 1 visible X reactions from 3 accounts; directional sample.
Ask a question below.
Published answers will appear here.