AI model gateway

One API. Hundreds of models. Built to ship.

Use an OpenAI-compatible interface to reach models across leading providers. Pay as you go with no subscription or seat fee.

Get an API key View models
  • OpenAI SDK compatible
  • Published model pricing
  • $5 minimum top-up
POST /v1/chat/completions
curl https://modelgates.ai/api/v1/chat/completions \
  -H "Authorization: Bearer $MODELGATES_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "anthropic/claude-sonnet-4",
    "messages": [{ "role": "user", "content": "Hello!" }]
  }'
OpenAI-compatiblebaseURL → modelgates.ai/api/v1
fact 01 · Catalog models
Hundreds
fact 02 · OpenAI-compatible API
1
fact 03 · Minimum top-up
$5
fact 04 · Chat requests per API key
10K/min

Models across leading providers

OpenAIAnthropicGoogleMetaMistralCohereDeepSeekxAIPerplexityTogether AIGroqFireworks AI
Core outcomes

A shorter path from model choice to production request

Compatibility, visible usage, and operational controls stay together behind one gateway.

Compatibility

Keep the interface you already use

Point an OpenAI-compatible client at one base URL and keep the familiar chat-completions request shape.

Usage-based billing

Pay for requests, not seats

Fund a prepaid balance from $5, see model pricing before you choose, and pay only for measured usage.

Operations

Control production traffic

Use separate API keys, enforced per-key concurrency limits, activity records, and optional auto top-up as workloads grow.

ModelGates / Adamant

Put policy between a model and its answer.

Adamant applies programmable governance to a model response and returns an auditable record of the rules, evidence, and decision alongside the result.

  • Policy-aware output
  • Two capability levels
Explore Adamant
governance receipt complete
Policy
workspace/default
Evidence
attached
Decision
passed
01 · PanelModels answer in parallel
02 · AnalysisAgreement and blind spots surfaced
03 · FusionOne synthesized response returned
ModelGates / Fusion

See how several models reach one answer.

Fusion runs a configurable panel in parallel, surfaces consensus and contradictions, then returns a synthesized answer with the intermediate analysis visible.

model: modelgates/fusion

Explore Fusion
Setup path

Three steps to your first request

Create an account, fund the balance, then generate a key and call the API.

Step 01

Create your account

Sign up with email or OAuth, then open the workspace that will hold your API keys.

Step 02

Fund the balance

Add at least $5 to the prepaid USD balance. Optional auto top-up can be configured later.

Step 03

Create a key and request

Create an API key, set the ModelGates base URL, and send an OpenAI-compatible request.

Pricing

Fund once. Pay for the usage you generate.

ModelGates uses a prepaid USD balance with per-model pricing and no subscription or seat fee.

Minimum top-up

$5

Add funds to a prepaid balance, then use them across the catalog. Auto top-up is optional and can be changed in credit settings.

Create an account

How billing reads

A visible path from catalog price to balance.

  1. 01Published model price
  2. 02Measured usage
  3. 03Prepaid balance debit
Compare model prices
Production needs

Built around production needs

Practical capabilities for adopting the gateway, managing usage, and operating API traffic.

need / 01

Switch without rewriting

Keep the OpenAI-compatible request shape. Change the base URL and API key, then choose a ModelGates model ID.

need / 02

Control spend as usage grows

Review published model prices, fund a prepaid balance, inspect usage activity, and enable auto top-up only when you need it.

need / 03

Keep workloads moving

Separate workloads with API keys, enforced per-key concurrency limits, and balance controls that keep production traffic deliberate.

Your next request

Start with one key and one familiar API.

Create an account to generate an API key, or read the quickstart before you connect a workload.

Create an account Read the quickstart