All models
I
Inception
Mercury 2.5
Built for response speed at a low price
This model is not in the catalog right now.
See available modelsAbout the model
Mercury 2.5 is an Inception model that puts response speed first. On routine tasks it answers very quickly, supports reasoning for complex ones and calls tools (function-calling) for agent scenarios. Context is 260 thousand tokens, input is text. The low price combined with speed makes it convenient where the answer goes to the user in real time and latency is noticeable. Russian at a medium level.
Strengths
- Fast response
- Reasoning when needed
- Context of 260K tokens
- Tool calling (function-calling)
- Low cost
Where to use it
- Interactive in-product assistants
- Autocomplete and suggestions
- Chatbots with a live response
- High-volume processing of short requests
- Agent scenarios at volume