3 /
What's new in Swarms InterLink?
Swarms InterLink makes agent workflows faster, more connected, and more efficient | Hanami
3 /
What's new in Swarms InterLink?
Swarms InterLink makes agent workflows faster, more connected, and more efficient through shared communication channels, reduced model-call overhead, full parallel execution, and more.
Highlights
- Communication Channels: Agents and structures can now share messages across six channel types, including files, queues, databases, HTTP and email.
- Faster model calls: swarms[fast] cuts per-call overhead by over 50% and memory usage from 220 MB to 100 MB, with no changes needed.
- More concurrency: ConcurrentWorkflow and GraphWorkflow now run entire layers in parallel. A 100-node workflow drops from 2.04s to 0.21s, outperforming LangGraph's 1.44s.
- Decision models: Support expands to OpenAI's gpt-6-luna, Jev and Clef. Switching models is simple, and every call now reports its cost in dollars.
Learn more: https://www.swarms.ai/blog/swarms-v17-interlink-release
Reply to @swarms_corp: 2 /
Get your update now with:
pip3 install -U swarms
Learn more: https://www.swarms.ai/blog/swarms-v17-interlink-release