Skip to main content
Control and security · Model gateway

Apply permissions and limits
to every model call

Connect models from 23 providers through vendor APIs, cloud, gateways or self-hosting. For each model, set which users and groups can use it, a daily token limit and a unit price to manage cost.

Contact sales
Provider list in Model connections
Provider list in Model connectionsReal product screen
LLM gateway · Code gateway

Apply the same policies
to every model call

Connect models from 23 providers, across vendor APIs, cloud, gateways and self-hosting, to one gateway. Requests from coding tools get the same policies and logging.

Code gateway pipeline
1
AuthenticationDeveloper verified by issued key
2
Policy checkBlocked repositories, files and commands
3
GuardrailsSensitive data inspection
4
Allowed modelHandled by an administrator-approved model
5
Coding session logCall record and cost tracking
Claude CodeCodex CLIOpenCode+8

Policy violations are logged as warnings by default, and administrators can switch them to blocking. Set which users and groups can use each model, along with daily token limits and pricing, to manage cost.

23Model providers4Connection typesAccess, limits and pricing per model
Real product screen
Code gateway policies and allowed models
Discriminative modelsNEW

Ready for the newest
class of model

Jev, released by TypeSafe in September 2026, is a discriminative model: it generates no text and returns one of the defined options with a probability. Syncgate adds discriminative models as a model type, connects any model that exposes the same API, and applies them to guardrail AI evaluation, auto evaluation and the search reranker.

Model management · model types
ChatEmbeddingRerankerImageGuardNEWDiscriminative
System One API connectionsThey share one API, so switching providers is a change of connection.
JJevTypeSafeProvider API
JJevVercel AI GatewayGateway
JJevDigitalOceanCloud
OOpenJevCodivHosted
OOpenJevvLLMSelf-hosted
KKevOllama · vLLMSelf-hosted
3Decision typesYes/no · choice · graded score
3FeaturesGuardrail AI evaluation · auto evaluation · search reranker
6ConnectionsProvider API · gateway · cloud · self-hosted

Latest jev-1.13.0 · version pinning

Guardrail AI evaluationSample

A yes/no probability is compared with two thresholds to decide the action.

Yes/noDoes the content request or disclose a customer’s personal contact details?

Give me the customer Minsu Kim’s mobile number.

Block probability0.91
Block

Conventional AI evaluation returns only pass or block. Because a discriminative model returns a probability, requests between the two thresholds are not blocked but kept in the guardrail inspection record.

Pass
Log
Block
Log threshold 0.35
Block threshold 0.70
00.350.701

Three decision types

The input is read once and several decisions run at the same time. Every result comes with a probability, which each feature uses as a score from 0 to 1.

Q. How many days ahead do I request annual leave?A. Request annual leave in the HR system at least 3 days before. [1]
Graded scoreAnswer helpfulnessAuto evaluation
Very lowLowMediumHighVery high
3.4 / 4 → 0.85
ChoiceMost relevant documentSearch reranker
HR policy 5.20.74
Leave FAQ0.18
Benefits guide0.08
Yes/noContradiction with the sourceAuto evaluation
Yes0.04

Uses assigned per model

Connect external services and self-hosted models side by side, and assign the discriminative model each feature uses.

Guardrail AI evaluationAuto evaluationSearch reranker
JevTypeSafe · external service
OpenJevSelf-hosted · GPU
KevSelf-hosted · Ollama

A newly connected discriminative model cannot be selected in any feature until it is assigned a use.

Coming nextTool-call approvalTool selection
Yes/no · choice · graded scoreProbability thresholdsUses assigned per modelSystem One APISelf-hosted for closed networks
What changes

Before and after adoption

Before

Each department opens its own AI accounts, so costs and usage records are scattered.

After

Every model call is handled by the same gateway, so usage and estimated cost appear on one screen.

Before

Company data may be sent to models nobody approved.

After

Only providers and models an administrator has approved are allowed, per group and per user.

Before

External models cannot be used on an air-gapped network.

After

Connect models your company runs itself with Ollama, vLLM, LM Studio, NVIDIA NIM and others.

How it works

Four steps

01Connect providers

Choose a provider, enter a key or URL, and confirm the connection with a real call.

Connection test result in the model connection dialog
Key features

Model gateway features in detail

23 providers

Connect in four ways: vendor APIs, cloud, gateways and self-hosting.

Permissions per model

Set which users and groups can use each model.

Daily token limits

Set request and token limits for each user.

Unit prices and budgets

Enter a unit price for each model to calculate estimated cost, and set budgets.

Models per task

Choose chat, embedding, reranker, image and guardrail models separately.

Discriminative models

Discriminative models, which answer with one of a fixed set of options and its probability, are used for guardrails and evaluation.

Admin settings

Settings each company configures

  • Connect only approved providers and models.
  • Set user and group permissions and token limits for each model.
  • Set unit prices and budgets to manage cost.
23Model providers
4Connection types
Access, limits and pricing per model

Figures reflect the product as of October 2026.

Related features

Related pages

Code gatewayApplies company policy to developers' coding toolsGuardrailsMasks personal data and blocks risky requestsEvaluation and tracingScores answer quality and records how each request is processedInstallation and operationsInstall and update on company servers and air-gapped networks
FAQ

Frequently asked questions about the model gateway

Which providers can I connect?

You can connect 23 in all: providers such as OpenAI, Anthropic, Google, Amazon Bedrock and Azure, gateways such as OpenRouter and LiteLLM, and models you run yourself with Ollama, vLLM, LM Studio, NVIDIA NIM and others.

Can it be used on an air-gapped network?

Yes. Register your in-house model's address as an OpenAI-compatible endpoint. Everything from the connection check to chat has been verified in an environment with no internet connection.

What happens when a limit is exceeded?

The conversation that goes over the limit runs to the end, and the limit applies from the next conversation. The user sees a notice that the request limit has been exceeded.

See it with
your own documents

In a consultation, we connect a few of your documents so you can check the answer quality for yourself. Send us your questions and our team will get back to you.

Contact salesRequest security documentation
Product videosReal product screens · Example company ACME Logistics · 16 videos