Trade latency for cost, per request

Service Tiers

Pick Flex or Priority processing on supported OpenAI, Google, Fireworks and AWS Bedrock models with one service_tier parameter.

Get startedRead the docs
service_tier · price multiplier
  1. 1

    flex

    0.5×
  2. 2

    default

    1×
  3. 3

    priority

    per model
01

Flex at 0.5×

Half price for work that can wait.

02

Priority lane

Faster processing when it matters.

03

Never downgraded

Retries keep your requested tier.

04

Fair billing

Billed on the tier actually served.

Built for

  • Batch jobs

    Evals and backfills on Flex

  • User-facing paths

    Priority for live traffic

  • Mixed workloads

    Choose per request, not per account

Ask for priority
terminal
1curl https://api.llmgateway.io/v1/chat/completions \2  -H "Authorization: Bearer $LLM_GATEWAY_API_KEY" \3  -H "Content-Type: application/json" \4  -d '{5    "model": "google-vertex/gemini-3.1-pro-preview",6    "service_tier": "priority",7    "messages": [8      {9        "role": "user",10        "content": "Summarize this report"11      }12    ]13  }'
Read docs

More in routing & reliability

All features

Dynamic Routes

Routing logic, out of your code

Sessions

Sticky routing for whole conversations

Caching

Pay once for repeated work

Start using Service Tiers today

One API key. Every model. Live in minutes.

Start buildingBrowse models