How do we govern which AI models employees use?

Routing rules say which models may answer which requests - by data boundary, region, cost or team - and every request that broke a rule is in the audit trail.

Does the platform store our prompts?

Not by default. The trace keeps the decision, the cost and the model; the prompt text is stored only if you turn retention on for your workspace.

Can we prove to an auditor what happened?

Yes: the audit trail is hash-chained, exportable and summarised in the governance report, and the security page says plainly which controls are checkable today and which are in progress.

AI governance

Control where every AI request goes. Then prove it.

Model allow- and deny-lists, provider and region restrictions, data boundaries, PII handling, budgets and quotas per workspace, tenant and key, approval through a human queue, and a hash-chained audit trail - applied before any model is scored, visible in a governance report with a posture score. Hosted or self-hosted, Apache-2.0.

Get a free API key Route this request in the playground

Free plan, no card. Open source under Apache-2.0; self-hosting is free forever.

From the benchmark

100%

of requests traced, attributed and auditable

Every metered request carries its key, member, tenant, target, tokens, cost and outcome; the governance report lists every control in force with the findings behind it. Controls you configure for regulated workloads - not a certification claim.

The whole benchmark, with the cases the router lost

The problem

A policy document does not stop a request.

Most AI governance is a page that says which models may be used, where data may go and who may spend what. Nothing reads it at request time. The platform team is asked to enforce it and has no place in the path of a request to do so - until the router is that place.

Allowed models and regions are a document, not a control
Spend is known per vendor after the invoice, not per team before the call
An auditor's question - where did this request go, and why - has no record to answer it

How the router answers it

Three things it does for every request.

Constraints before the score
Data boundary, region, tenant, cost and latency ceilings, allowed and denied targets, PII handling and payload logging apply to every key before any candidate is scored. A target that fails is not a candidate, and the trace says which rule removed it.
Budgets that refuse
Daily and monthly budgets per workspace, tenant and key are ceilings, not alerts: once reached, the request is refused with a reason the caller can read. Per-key scopes, rate limits and expiry bound what one credential can do.
On the record
A hash-chained audit trail of decisions and outcomes, a governance report with every control in force and a posture score, and - when judgement is needed - a human queue as a routing target with approval in the loop.

Watch one decision

Illustrative: the lowest cost that still clears the quality the task needs. The live quote below is the real one.

A real request

The workspace policy, as the governance page writes it.

Not a prompt and not a weight: a request that would break it is refused or re-routed before a model is called, and the audit trail records the rule.

Route it in the playground and read the trace

Exampleyaml

data_boundary: private        # never a public endpoint
region: eu                    # EU targets only
contains_pii: true            # treat every request as if it may carry PII
deny_targets: [llm-frontier]  # never this one, whatever the score
max_cost_per_1k: 0.01
daily_budget_usd: 250
log_payloads: false           # prompts and answers are not kept

Where this stands next to the alternatives, and how to go deeper.

Install

curl -fsSLO https://opensmartroute.ai/download/helm && helm install router opensmartroute-*.tgz

The enterprise architecture guide

Or no install at all: the hosted API answers a plain curl with the decision, and the browser extension shows the price and the router's pick under the composer of the AI chat sites you already use.

For the business

What it does for the risk register.

A trace with every answer
Every decision explained
Which model answered, why it won, what it cost and what the alternatives were - kept for the retention period you set.
Open a trace
Redacted before it leaves
Personal data stays home
Names, numbers and identifiers are removed on your side before the request reaches a provider; prompts are not stored by default.
The security page
Hash-chained audit trail
An audit you can hand over
Decisions, policy changes and sign-ins land in a tamper-evident log, and the governance report summarises them for the people who ask.
What is checkable today

How do we govern which AI models employees use?: Routing rules say which models may answer which requests - by data boundary, region, cost or team - and every request that broke a rule is in the audit trail.
Does the platform store our prompts?: Not by default. The trace keeps the decision, the cost and the model; the prompt text is stored only if you turn retention on for your workspace.
Can we prove to an auditor what happened?: Yes: the audit trail is hash-chained, exportable and summarised in the governance report, and the security page says plainly which controls are checkable today and which are in progress.

Route your first request in under a minute.

OpenSmartRoute is open source and the free plan keeps the full trace of every decision. Create a workspace, mint a key and send the request above.

Create a free account Read the benchmark

Free plan, no card. Fifteen thousand decisions a month with the full trace.

Control where every AI request goes. Then prove it.

data_boundary: private # never a public endpoint region: eu # EU targets only contains_pii: true # treat every request as if it may carry PII deny_targets: [llm-frontier] # never this one, whatever the score max_cost_per_1k: 0.01 daily_budget_usd: 250 log_payloads: false # prompts and answers are not kept

Control where every AI request goes. Then prove it.

A policy document does not stop a request.

Three things it does for every request.

Constraints before the score

Budgets that refuse

On the record

The workspace policy, as the governance page writes it.

Where this stands next to the alternatives, and how to go deeper.

What it does for the risk register.

Every decision explained

Personal data stays home

An audit you can hand over

Route your first request in under a minute.

Control where every AI request goes. Then prove it.

A policy document does not stop a request.

Three things it does for every request.

Constraints before the score

Budgets that refuse

On the record

The workspace policy, as the governance page writes it.

Where this stands next to the alternatives, and how to go deeper.

What it does for the risk register.

Every decision explained

Personal data stays home

An audit you can hand over

Route your first request in under a minute.