Introducing FloTorch.AI v1.3 – Smarter Workflows, Deeper Integrations, and More Control

At FloTorch.AI, we’re building the future of agentic AI workflows — faster, simpler, and more scalable than ever before. With every release, our mission is to empower teams across the globe to orchestrate intelligent systems with less friction and more flexibility.

And today, we’re thrilled to announce FloTorch v1.3 (released Sept 19, 2025) — packed with powerful new features, smarter routing, expanded integrations, and enhanced analytics to help you take your AI projects to the next level.

What’s New in FloTorch 1.3

🖱️ Build Workflows with Drag & Drop

Say goodbye to complex orchestration. With our no-code Drag & Drop Prompt Sequencing, you can visually design agentic workflows — no engineering support required.

⚡ Execute with Runtime Power

Whether you prefer no-code simplicity or direct API calls, the new Execution Runtime makes running entire workflows seamless through the SDK.

🧠 Smarter Memory with Agentic Core

Choose the right memory for your agents. With Agentic Core Memory now available alongside Mem0, you get flexibility today and a future-ready foundation for tomorrow.

⏱️ Smarter Routing, Less Waste

Time-Based Routing → Automatically switch between models based on demand cycles (fast LLMs at peak, cost-efficient ones during low traffic).
Keyword-Based Routing → Route queries to domain-specific LLMs (finance, healthcare, research) with simple keyword triggers.

🔗 Expanded Integrations

Vector Databases: LanceDB, PineconeDB, ChromaDB, OpenAI Search
LLM Providers: Latest models from Groq, Together AI, and Fireworks AI — all supported via OpenAI-compatible layers

💰 Save More with Cache Insights

Our dashboard now highlights exactly how FloTorch caching reduces token usage — helping your team optimize costs while maintaining performance.

🔧 Enhancements That Matter

Memory & Tools are now first-class features in the Console — more visibility, more control.
Workspace User Access Roles to keep governance in check.
LLM Listing Page with models sourced from models.dev.
Code Editor Support for Cline & roo code.
GCP Integration — deploy with confidence on Google Cloud.