Warden Announces Full Migration to Venice.AI

TL;DR 

We’ve moved all of Warden’s AI workloads to Venice.AI. Since launch, Warden’s proprietary agents have consumed nearly 1 trillion tokens and now handle 1.5M+ daily chats—all of which will run on Venice models. Warden’s upcoming Agent Network will provide onchain and offchain agent builders with access to Venice’s private, uncensored AI. 

Why We’re All-In on Venice

Over the last year, Warden grew from a cutting edge co-pilot into the beginnings of an Everything App for crypto. That scale forced a choice: compromise on performance with centralized AI platforms, or standardize on an AI stack that aligns with our values.

We chose Venice.AI. Here’s why:

• Uncensored models: AI models that are impervious to bias where developers control behavior – no surprise content restrictions, no platform handbrakes.

• Private, user-oriented architecture: Venice is the only AI inference provider operating at scale that doesn't store user prompts and chats.

• Venice  scales seamlessly: as Warden’s throughput increased, Venice scaled alongside us through VVV staking for inference.

What Changes Today

• Full migration in-flight: Warden AI traffic will run on Venice models. Your daily experience stays fast and familiar—just more private and resilient.

• Scale, proven: Since launch, Warden’s proprietary agents have consumed ~1T tokens. We now average 1.7M+ daily chats, all moving to Venice.

• Deeper integration: Venice models already power Warden App, and they’re rolling into Warden Studio as first-class models for developers. Agent developers can invoke Venice models via Warden’s infrastructure.

A Word from the Founders

“By migrating fully to Venice, we’re standardizing on Venice’s uncensored, verifiable intelligence that scales with our users and builders. This is the backbone of the Warden Ecosystem powering our agents that can reason, execute, and coordinate across chains.”—Josh Goodbody, Founder, Warden Labs

“Most AI platforms treat developers like children who can't be trusted with raw machine intelligence. With Venice’s uncensored models now fully integrated into Warden's stack, we’re giving developers access to truly limitless AI that follows their instructions instead of opaque corporate content policies.”—Erik Voorhees, Founder & CEO, Venice.AI

What Developers Get

• Venice inside Warden Studio: Use Venice models as plugins—wire them into routing, planning, tool invocation, and image generation. Build agents that are variable and have access to millions of users on Warden’s Agent Network.

• No deplatforming risk: Uncensored models reduce the chance that a policy update breaks your product.

FAQs

Is anything changing in the Warden App UI?

Not immediately. The change is under the hood: all model traffic will use Venice. You should notice stable performance.

Can I build agents that call Venice from my own application?

Yes. That’s precisely what our integration enables: contract and agent-level access to Venice models via Warden’s infra, with verifiable outputs. 

What about content restrictions?

Venice provides uncensored models, minimizing deplatforming risk and preserving developer control. 

When do Venice plugins land in Warden Studio?

Initial models are available today, with more soon. Venice already powers the Warden App; Studio integrations are rolling out with docs and examples. 

Questions or looking to co-build? Ping us and we’ll get you into the next wave of Warden Studio access.