AIBaseten launches web search preview for open models
Baseten says its new Hosted Tools and Grounded Inference offer 15% lower latency than client-side execution, with no extra vendor key required and no orchestration.
TLDR
Baseten announced a preview of Hosted Tools and Grounded Inference, which it says bring real-time web search server-side to open models running on its platform through a single configuration. The company claims 15% lower latency than client-side execution, with no extra vendor key required and no orchestration. It names four web search partners for the preview: @ExaAILabs, @KeenableAI, @p0 and @youdotcom.
Combined views
211
1 Source, first seen 1d ago
Baseten launches web search preview for open models
Baseten says its new Hosted Tools and Grounded Inference offer 15% lower latency than client-side execution, with no extra vendor key required and no orchestration.
TLDR
Baseten announced a preview of Hosted Tools and Grounded Inference, which it says bring real-time web search server-side to open models running on its platform through a single configuration. The company claims 15% lower latency than client-side execution, with no extra vendor key required and no orchestration. It names four web search partners for the preview: @ExaAILabs, @KeenableAI, @p0 and @youdotcom.