MyAISpeed GitHub Action

Test connection latency to major AI providers directly from your CI/CD pipeline.

Quick Start

- uses: myAIspeed/action@v1id: speed
- run: echo "Fastest: ${{ steps.speed.outputs.fastest-provider }} (${{ steps.speed.outputs.fastest-latency }}ms)"

Usage

- uses: myAIspeed/action@v1id: speedwith:
providers: 'all'# Comma-separated list or 'all'fail-threshold: '0'# Fail if any provider exceeds this ms (0 = never fail)json-output: 'false'# Log results as JSON

Inputs

Input	Description	Default
providers	Comma-separated list of provider slugs to test, or all	all
fail-threshold	Fail the step if any provider latency exceeds this value in ms. Set to 0 to never fail.	0
json-output	Set to true to include JSON-formatted results in the log output	false

Outputs

Output	Description
results	JSON array of all test results
fastest-provider	Slug of the fastest responding provider
fastest-latency	Latency of the fastest provider in milliseconds

Examples

Fail if latency exceeds 500ms

steps:
- uses: myAIspeed/action@v1id: speedwith:
fail-threshold: '500'

Save results as an artifact

steps:
- uses: myAIspeed/action@v1id: speedwith:
json-output: 'true'
- run: echo '${{ steps.speed.outputs.results }}' > latency-results.json
- uses: actions/upload-artifact@v4with:
name: ai-latency-resultspath: latency-results.json

Test specific providers

steps:
- uses: myAIspeed/action@v1id: speedwith:
providers: 'openai,anthropic,google'

About

Built by Reflect Memory, Inc. — measure and monitor AI provider performance from anywhere.

Visit myaispeed.com to learn more.