Platform
MCP Server Build
A production MCP server for one of your systems, with up to 8 tools, secure sign-in, action logging and tests in AI…
Platform
AI Gateway
Every AI request from every workflow passes through one gateway, with keys, rate limits and logs in one place.
When every application calls AI providers directly, keys end up scattered, limits are set in five places and nobody can say what each workflow costs. AI Gateway routes all those requests through one point you control, with keys, rate limits and logging handled centrally. It works across providers. Your team keeps the configuration in version control.
- One place for every provider key, limit and log
- Usage, latency and cost visible for each workflow
- Switch or add model providers without touching every application
$4,500
20–30 working days
Optional maintenance $249/mo
What it is
AI Gateway is a build that puts a single integration point in front of every model provider your team uses, handling API keys, rate limits and usage logging centrally instead of scattered across each application. It works with multiple providers and reports usage, latency and cost for each workflow.
Why businesses buy it
As AI use grows, so does the mess behind it. API keys sit in several codebases, nobody owns rate limits, and the monthly bill arrives with no breakdown by workflow. That is a security risk and a budgeting problem. After this build, every request passes through one gateway, so keys are held in one place, limits are enforced consistently and you can see exactly where usage goes. The Scope Card lists the applications, providers and controls in writing before you order, so the work is defined and the outcome is clear.
Problems this solves
API keys for AI providers are scattered across every application
The AI Gateway puts one integration point in front of your model providers. API keys are handled centrally instead of being copied into each of up to three applications.
One busy app hits rate limits and breaks the others
Rate limits are managed at the gateway, with up to five controls configured from your policies. Each workflow's use is visible, so one app does not quietly starve the rest.
Nobody can say which workflow drives AI cost or delay
The gateway logs usage centrally and reports usage, latency and cost per workflow across providers. The admin console or written report shows where time and spend go.
A good fit if
You're running AI workflows against more than one provider and each one currently manages its own keys and limits separately.
You want usage, latency and cost visible per workflow, not buried inside each application's own logging.
You have a named technical owner who can supply cloud or hosting access and review the configuration before launch.
You want provider swaps or fallback routing handled centrally, without changing code in every application that calls a model.
Not the right fit if
You specifically need the decision logic for which model handles a request, not the full integration layer. See AI Model Router.
Your priority is versioning and testing prompts, not routing requests between providers. See AI Prompt Management.
You need agents to retain context across sessions, not a routing layer for API calls. See AI Agent Memory System.
Your workflows need structured access to your own documents, not provider routing. See AI Agent Knowledge Layer.
How it works
- We map every application and workflow currently calling a model provider directly, so nothing is missed when traffic moves behind the gateway.
- API keys for each provider, such as OpenAI, Google Gemini, AWS Bedrock or Azure OpenAI, are consolidated into one managed layer.
- Rate limits and usage quotas are configured per workflow, so one application can't exhaust the budget or throughput meant for another.
- Every request is logged with latency, cost and outcome, exportable to a tool such as Langfuse or Grafana for monitoring.
- Applications are switched over to call the gateway instead of providers directly, tested against real traffic before the cutover completes.
- A control test log documents each configured control, reviewed with you before the gateway goes fully live.
What you provide within 72 hours
Your portal lists these the moment you order. Delivery starts the working day after everything is in and your access is checked.
- Written authorization confirming which applications and provider accounts are in scope for the gateway to manage.
- A list of the AI applications and systems that currently call a model provider directly.
- Your existing usage policies, such as rate limits or budget caps you already enforce informally.
- Cloud or hosting access needed to deploy the gateway alongside your existing infrastructure.
- A named technical owner who can review the routing configuration and sign off before cutover.
- Admin access and API keys for every connected provider, shared by adding our team as a user or creating a key you control, never by email.
- Your business information (company basics, hours and brand voice), shared during intake, and one decision-maker who can approve the final configuration.
The 72-hour rule. Send everything on your checklist, including access, within 72 hours of ordering so your build starts on time.
Timeline
Start
Intake verified, applications and provider accounts in scope confirmed.
Weeks 1–2
Gateway built with provider keys, rate limits and logging configured.
Week 3
Applications connected to the gateway and tested against real traffic.
Weeks 3–4
Control test log produced and shared with your technical owner.
Week 4
Your review and two rounds of changes applied to the configuration.
Go-live
Live, recorded handoff session, 30 days of support begins.
| Option | Delivery time | Price |
|---|---|---|
| Standard | 20–30 working days | $4,500 |
| Express | 15–25 working days | $5,300 |
| Rush | 10–20 working days | $6,100 |
What you receive at handover
- AI Gateway configured and deployed in front of your model providers, covering up to three applications
- Central key handling, rate limits and up to five controls built from your policies
- Usage, latency and cost reporting per workflow in an admin console or written report
- A test report with a control test log for the agreed scenarios
- An admin guide and a recorded 60-minute handoff session
- 30 days of post-launch support
Expected benefits and how to measure them
- Can make provider keys easier to manage and rotate.
- Helps you see which workflows drive AI cost and latency.
- Makes it simpler to add or switch a provider later.
- AI cost per workflow
- Read cost per workflow from the gateway report each month and compare with provider invoices.
- Rate-limit errors
- Count rate-limit failures from app logs before launch and from the gateway after.
- Places where API keys are stored
- Count them in an audit before launch and again after handover.
Before you start: List every app that calls an AI provider and where its keys are stored. Note last month's provider spend and any rate-limit incidents.
These are things to track, not promised results. Outcomes depend on your data, your team and your market.
Running costs
The build price is one-time. Running the Gateway afterward costs roughly $20 to $300 a month in your own accounts, covering hosting, logging and monitoring tools, and it moves with how much request volume passes through it.
AI model usage during the build and testing is billed to your own accounts too, so every provider relationship stays yours, not ours.
Ask about running costsCustom requirements
Need more than three systems in scope, more than five controls configured, or a second admin console? Add a note at checkout and you get a written Change Quote, only starting once you accept it. Larger or bespoke projects go through the Custom AI request form instead.
Request a custom versionKeep it running
Monthly maintenance covers monitoring, fixes and change hours, enough to add a provider or adjust a rate limit. It starts once the 30 days of support end and can be cancelled any month.
$249/mo (the minimum fee, includes 2 change hours)
Optional. Starts after the post-launch support period. Cancel any month.
How maintenance worksHow this compares
| Do it yourself with a SaaS tool | Hire a custom agency | Upmark Labs | |
|---|---|---|---|
| Price | An API gateway platform charges a monthly fee per request volume or per seat | A systems integrator quotes per project after a discovery phase | Ours is one fixed price, shown on this page before you buy |
| Scope | A gateway platform is infrastructure your engineers configure and maintain | An integrator agrees scope during the project | Ours is written on the Scope Card before purchase |
| Provider coverage | A gateway platform supports the providers it ships integrations for | An integrator build depends on who is staffed on it | Ours covers OpenAI, Gemini, Bedrock and Azure OpenAI as standard |
| Delivery | A gateway platform is live once your team finishes integrating it | An integrator usually takes several weeks to staff and start | Ours has a fixed delivery time shown before you buy, with faster options |
Questions about this product
How much does AI Gateway cost?
The fixed price for each delivery speed is shown in the price and delivery panel on this page before you order. It covers everything on the Scope Card: consolidating provider keys for up to three applications, configuring up to five controls such as rate limits and logging, and the control test log you review before cutover to the gateway.
Which model providers does AI Gateway support?
The standard build covers up to three systems or AI applications routed through providers such as OpenAI, Google Gemini, AWS Bedrock and Azure OpenAI. You choose which providers and applications are in scope during intake, and each connection is tested against real traffic before the gateway takes over routing.
What does AI Gateway cost to run each month?
Roughly $20 to $300 a month in your own accounts, covering hosting, logging and monitoring tools such as Langfuse or Grafana. You pay providers directly, so there's no markup from us on request volume. Light traffic sits toward the bottom of that range, high-volume production workflows toward the top.
How is AI Gateway different from AI Model Router?
AI Gateway is the full integration layer: keys, rate limits and logging for every provider in one place. AI Model Router is narrower and focused on the decision logic for which model handles a given request, by cost, capability or latency. The two can work together, with the Router's logic running inside the Gateway.
Do we need to change our application code to use AI Gateway?
Applications are switched to call the gateway's single endpoint instead of each provider directly, which is a configuration change rather than a rewrite of your AI logic itself. We test each application against real traffic during the build, and cutover happens once your technical owner confirms it behaves as expected.
Can we add more systems to AI Gateway later?
Yes. The standard build covers up to three systems or AI applications, and adding a fourth or beyond is quoted as a Change Quote once you are ready to expand. We suggest starting with your highest-traffic workflows, confirming routing and logging behave correctly, then bringing further applications under the gateway.
What do we need to provide before AI Gateway can launch?
Written authorization for the applications and provider accounts in scope, a list of the AI applications currently calling providers directly, your existing usage policies, and cloud or hosting access. We also need a named technical owner and admin access shared by adding our team as a user or creating a key you control, never sent by email.
Can I cancel maintenance for AI Gateway?
Yes, any month. Maintenance starts after the 30 days of post-launch support end and runs month to month with no cancellation fee. It includes a monthly allowance of change hours that scales with the build, and unused hours in a given month don't carry over to the next one.
Stop managing provider keys and limits application by application.
Order AI Gateway and every AI request routes through one managed layer on a fixed delivery date. The scope above is what you get, in writing, before you pay.
Works well with
Platform
AI Agent Orchestration System
Coordinates multiple agents, tools and handoffs in one runtime.
