OpenRouter Launches Tech to Slash AI Multi-Turn Conversation Costs
As AI conversation technology evolves, OpenRouter has launched two new features—Prompt Caching and Sticky Routing—designed to cut token costs in multi-turn agent conversations. This innovation allows developers to reduce expenses during extended interactions with agents, encouraging wider adoption of agent applications.
In multi-turn conversations, agents frequently resend the same system prompts, tool definitions, and strategy instructions. Even if the user only changes one input across six rounds, each repeated prompt incurs charges. Prompt Caching lets OpenRouter retrieve cached data, bypassing the cost of re-sending identical prompts. Depending on the provider, cached reads cost 0.1 to 0.5 times the standard input price, significantly lowering expenses. For instance, on Claude Sonnet 4.6, cached reads are just $0.30 per million tokens, compared to $3.00 for normal input.

To maintain cache effectiveness across multiple turns, OpenRouter's Sticky Routing sends subsequent requests to the same provider that holds the hot cache, enabling developers to consistently enjoy low-cost cache reads throughout the conversation.
However, to take full advantage of this innovation, developers must keep the same session throughout. For optimal caching, OpenRouter recommends setting a session_id, which activates Sticky Routing immediately after the first successful request—not only after a cache hit. Developers can also verify cache usage by checking the response data.
As AI technology progresses, OpenRouter's latest feature not only gives developers a powerful cost-control tool but also opens new possibilities for agent applications.
Related article
Stripe to Acquire OpenRouter as AI Model Routing Gains Momentum
Stripe has agreed to acquire OpenRouter, an AI model-routing platform that gives developers access to hundreds of models through a single interface. The deal adds model selection and routing to Stripe’s existing work around AI usage and token-based b
Former OpenAI Researcher Barrett Zoph Returns to Google to Support Gemini Development
Barret Zoph, who previously abruptly left and joined OpenAI due to a disagreement with Thinking Machines Lab CEO Miriam Mularati, has once again changed employers. Alphabet's Google announced that Zoph will return as Vice President of Research, bring
How Recursive Intelligence Raised $335M at $4B Valuation in 4 Months
The co-founders of startup Ricursive Intelligence seemed destined to be co-founders.Anna Goldie, CEO, and Azalia Mirhoseini, CTO, are so well-known in the AI community that they were among those AI engineers who “got those weird emails from Zuckerber
Related Special Topic Recommendations
Comments (0)
0/500
As AI conversation technology evolves, OpenRouter has launched two new features—Prompt Caching and Sticky Routing—designed to cut token costs in multi-turn agent conversations. This innovation allows developers to reduce expenses during extended interactions with agents, encouraging wider adoption of agent applications.
In multi-turn conversations, agents frequently resend the same system prompts, tool definitions, and strategy instructions. Even if the user only changes one input across six rounds, each repeated prompt incurs charges. Prompt Caching lets OpenRouter retrieve cached data, bypassing the cost of re-sending identical prompts. Depending on the provider, cached reads cost 0.1 to 0.5 times the standard input price, significantly lowering expenses. For instance, on Claude Sonnet 4.6, cached reads are just $0.30 per million tokens, compared to $3.00 for normal input.

To maintain cache effectiveness across multiple turns, OpenRouter's Sticky Routing sends subsequent requests to the same provider that holds the hot cache, enabling developers to consistently enjoy low-cost cache reads throughout the conversation.
However, to take full advantage of this innovation, developers must keep the same session throughout. For optimal caching, OpenRouter recommends setting a session_id, which activates Sticky Routing immediately after the first successful request—not only after a cache hit. Developers can also verify cache usage by checking the response data.
As AI technology progresses, OpenRouter's latest feature not only gives developers a powerful cost-control tool but also opens new possibilities for agent applications.
Stripe to Acquire OpenRouter as AI Model Routing Gains Momentum
Stripe has agreed to acquire OpenRouter, an AI model-routing platform that gives developers access to hundreds of models through a single interface. The deal adds model selection and routing to Stripe’s existing work around AI usage and token-based b
Former OpenAI Researcher Barrett Zoph Returns to Google to Support Gemini Development
Barret Zoph, who previously abruptly left and joined OpenAI due to a disagreement with Thinking Machines Lab CEO Miriam Mularati, has once again changed employers. Alphabet's Google announced that Zoph will return as Vice President of Research, bring
How Recursive Intelligence Raised $335M at $4B Valuation in 4 Months
The co-founders of startup Ricursive Intelligence seemed destined to be co-founders.Anna Goldie, CEO, and Azalia Mirhoseini, CTO, are so well-known in the AI community that they were among those AI engineers who “got those weird emails from Zuckerber





Home






