tmf test — Health Check

Description

tmf test runs a connectivity test against a specified LLM provider, measuring latency and throughput to verify the configuration is correct and the service is reachable.

Syntax

tmf test <provider> [options]

• <provider> — Provider name, e.g. deepseek, openai

Options

Option	Description
-k, --key <api-key>	API Key
-m, --model <model>	Model to use for the test
-p, --prompt <text>	Custom test prompt
-v, --verbose	Verbose output with detailed results

Examples

# Quick connectivity test for DeepSeek
tmf test deepseek

Example output (normal mode):

🔍 Testing deepseek (deepseek-chat)...
   Endpoint: https://api.deepseek.com/v1

Test complete  Latency 1200ms  Token usage 42  Throughput 28.0 token/s

# Test OpenAI with a specific model and custom prompt
tmf test openai --model gpt-4o-mini -p "Say hello"

# Verbose output with full details
tmf test openai --verbose

Example output (verbose mode):

🔍 Testing openai (gpt-4o)...
   Endpoint: https://api.openai.com/v1

   Prompt: Hello, please introduce yourself in one sentence.

   First token arrival: 800ms
   Token usage: 28 (prompt: 12, completion: 16)
   Generation time: 1.5s
   Throughput: 10.7 token/s

# Test with an explicit API key
tmf test deepseek --key sk-xxxxxxxx