TrueFoundry (@truefoundry) / X

TrueFoundry

336 posts

TrueFoundry

@truefoundry

TrueFoundry is a unified platform with an enterprise-grade AI Gateway - combining LLM, MCP, and Agent Gateway to securely manage, route, and govern AI workloads

San Francisco

Joined April 2022

Pinned
TrueFoundry
@truefoundry
Jun 8
The worst advice for enterprise AI agents? Start small. "Start small" sounds safe. In reality, it often leaves enterprises with 100+ agent pilots and no scalable AI strategy. The winners do the opposite: → Start with a high-value problem → Build the foundation → Scale from
00:00
TrueFoundry
@truefoundry
2h
@truefoundry now integrates with @GraySwanAI . Teams building LLM applications and agents can now enforce Gray Swan Cygnal security policies through the TrueFoundry AI Gateway. Detect prompt injections, jailbreaks, harmful content, and custom policy violations inline across
Made with AI
TrueFoundry
@truefoundry
Jul 24
@truefoundry now integrates with @ringg_ai . Teams building AI voice applications and agents can now route speech-to-text (STT) requests through the TrueFoundry AI Gateway to Ringg AI using a unified OpenAI-compatible interface without changing application code. Organizations
Made with AI
TrueFoundry
@truefoundry
Jul 24
OpenAI's GPT-5.6 pricing quietly added a cost to caching: cache writes now run 1.25x normal input price. We modeled the break-even point, and it's the same across Sol, Terra, and Luna: writes can be up to 78.3% of requests before caching stops saving you money. Read the full
TrueFoundry
@truefoundry
Jul 24
We ran GLM 5.2 vs Claude Opus 4.8 on tasks built to expose the gap. Quality: tied. The real gap wasn't intelligence, it was tokenizers: a 1M-token doc hard-failed Opus's context window but GLM handled it fine. GLM's also ~4x cheaper, with spiky latency on hard tasks. Read the
See @truefoundry's full profile
Sign up
Log in

See @truefoundry's full profile

See @truefoundry's full profile