Docs
CommunityJev Integrations & Adapters
MLflow AI Gateway Jev Providers
MLflow's AI Gateway lists TypeSafe Jev among its evaluation providers: a typesafe passthrough at /gateway/typesafe/v1/systemone forwards raw System One requests using the server-side TypeSafe key, separate from MLflow judges.
- Format
- Docs
- Reading time
- 2 min
- Published by
- Community
- Author
- MLflow
- Last verified
- Oct 2026
- Languages
- Python
- Capabilities
- Self-hostedRequires API key
Highlights
- The TypeSafe passthrough is POST /systemone under http://localhost:5000/gateway/typesafe/v1, forwarding to https://api.typesafe.ai/v1 by default.
- OpenRouter's Jev decision models share the /gateway/typesafe/v1/systemone route; typesafe/jev-router is a chat model on normal chat routes.
- The gateway endpoint name goes in the request body's model field, and the gateway attaches its server-side TypeSafe API key.
- Jev routes support only structured System One evaluation, not the gateway's chat, embeddings, OpenAI-compatible, or streaming APIs.
- The curl example asks a noul whether state.outputs answers the question in state.inputs, under the model name jev-evaluator.
Watch out
Vendor documentation for MLflow's AI Gateway, which runs on localhost with LiteLLM available only as an optional extra. It shows no Jev pricing, licence, or rate limits, and it warns that the raw System One route is not a chat endpoint and must not be passed to make_judge as a gateway model.
Questions about this entry
Short answers drawn from the linked project or page.
As evaluation providers: TypeSafe and OpenRouter Jev decision models both use the /gateway/typesafe/v1/systemone passthrough, which forwards structured System One evaluation requests with the gateway's server-side TypeSafe API key.
POST to http://localhost:5000/gateway/typesafe/v1/systemone with the endpoint name in the model field. Requests go to https://api.typesafe.ai/v1 by default, and a proxy or another System One host can be set through api_base.
No. TypeSafe and OpenRouter Jev decision endpoints support only structured System One evaluation requests; the gateway's chat, embeddings, OpenAI-compatible, and streaming APIs are not available through this route.
TypeSafe gateway endpoints are raw System One endpoints, separate from MLflow judges created with make_judge(model='typesafe:/...'). Do not pass a TypeSafe gateway endpoint as model='gateway:/...' to make_judge.
The gateway endpoint name goes in the request body's model field, and the gateway forwards the request to the configured TypeSafe model. The example uses jev-evaluator for an answer-quality noul over state.inputs and state.outputs.
No. MLflow AI Gateway supports many providers natively; LiteLLM can optionally be installed to extend the gateway to providers not listed, but it is not a dependency and Jev support does not require it.
No. The page lists TypeSafe and OpenRouter as evaluation providers with the /systemone passthrough and gives curl and SDK examples, but includes no pricing, model aliases, or rate limits and no benchmark information.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
X: Jev 1.14 is now on OpenRouter
X: Jev lands on the Vercel AI Gateway
X: Jev as an LLM guardrail inside Agentgateway
YouTube: Greg Isenberg and Ryan Vogel on the businesses Jev unlocks, and a path in through the Vercel AI Gateway.
More Jev resources like this
Pydantic's AI Gateway forwards Jev natively as a bring-your-own-key TypeSafe provider, so one gateway key covers spend views, limits, and input guardrails while TypeSafeModel turns each output field into a question.
Jev Integrations & AdaptersArticle#community#pydantic#gateway
Community
LakeSail built five Jev functions into Sail's SQL, so Spark SQL and DataFrame queries can ask typed questions about every row, with calls overlapping in the Rust engine behind one in-flight limit per worker.
Jev Integrations & AdaptersArticle#community#sql#rust
Community
Autonomous agent platform with 187,504 stars for building and running agents from plain English or a visual builder. Its platform features first-class TypeSafe Jev decision blocks for typed routing, filtering, scoring and confidence-gated next-action dispatching.
Jev Integrations & Adapters#community#autonomous-agents#platform
Community
Significant-Gravitas/AutoGPTPython
Significant-Gravitas/AutoGPTPythonLiteLLM's proxy adds TypeSafe guardrail hooks and a Jev complexity classifier for policy judgments and request routing across 100+ model providers.
Jev Integrations & Adapters#community#python#gateway
Community
BerriAI/litellmPython
BerriAI/litellmPythonVolcengine's agent context database that unifies memory, knowledge RAG and skills under a viking:// virtual filesystem; it ships a Jev rerank client that scores candidate documents against api.typesafe.ai and treats the probability as relevance.
Jev Integrations & Adapters#community#agent-memory#rag
Community
volcengine/OpenVikingPython
volcengine/OpenVikingPythonA semantic cache for LLM queries, fully integrated with LangChain and llama_index. Zilliz's cache integrates TypeSafe Jev Noul checks to evaluate cache hit freshness and time-dependent query validity, cutting repeated requests and API cost.
Jev Integrations & Adapters#community#semantic-cache#langchain
Community
zilliztech/GPTCachePython
zilliztech/GPTCachePythonFrom the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.