Levanto says its guardrail model matches GPT-5 on AgentHarm and runs 5x faster
w1kke · x · 2026-07-28
Levanto Labs says its new guardrail model matches GPT-5 on AgentHarm while running 5x faster.
- The company argues that today’s guardrails force a tradeoff between intelligence and latency.
- Current classifier-based systems are fast, but they often lack full conversation context and fail on complex reasoning.
- LLM-based guardrails are smarter, but too slow for the hot path.
- Sage combines LLM and classifier traits and reportedly responds in about 200ms, or 5x faster than GPT-5 with reasoning set to minimal and 19x faster if GPT-5 uses low reasoning.
More from Safety
- Cursor is accused of uploading 63,106 files and 736 MB of source code — kristoph · 2026-07-28
- A report says OpenAI’s pre-release models already exposed internal deployment risks — ruthstarkman · 2026-07-28
- Black Hat side event says AI agents are shrinking the zero-day window to hours — jcran · 2026-07-28
- Meta signs EU AI transparency code but warns too many labels reduce clarity — HaktanSuren · 2026-07-28
- CTI Expert turns Claude into a cyber threat intelligence analyst with 74+ commands — tom_doerr · 2026-07-28
- Post-mortem says the HF/OpenAI incident was a test-environment failure, not a Skynet attack — deliprao · 2026-07-28