Predictive Models through an API
Server: BART API server.
Client: Me, on my phone.

Server: BART API server.
Client: Me, on the Transit App.

Server: BART API server.
Client: Me, on the Transit App.

Server: My CT-Predict API Server.
Client 1: My CT-Predict App.
. . .
Client 2: Kaiser’s EHR system.

Server: OpenAI API server.
Client: Me, on the ChatGPT App.

Server: Anthropic API server.
Client: Me, on the Claude.ai App.

Server: Anthropic API server.
Client: Me, at the terminal.

curl https://api.anthropic.com/v1/messages \
-H "x-api-key: $ANTHROPIC_API_KEY" \
-H "anthropic-version: 2023-06-01" \
-H "Content-Type: application/json" \
-d '{
"model": "claude-haiku-4-5-20251001",
"max_tokens": 100,
"messages": [
{
"role": "user",
"content": "The capital of France is"
}
]
}'
{
"model":"claude-haiku-4-5-20251001",
"id":"msg_012mMQuMNGPFXNZqookPvuu9",
"type":"message",
"role":"assistant",
"content":[{
"type":"text",
"text":"The capital of France is **Paris**."}],"stop_reason":"end_turn",
"stop_sequence":null,
"usage":{
"input_tokens":12,
"cache_creation_input_tokens":0,"cache_read_input_tokens":0,
"cache_creation":{
"ephemeral_5m_input_tokens":0,"ephemeral_1h_input_tokens":0},
"output_tokens":11,
"service_tier":"standard"
}
}What it does:
I’ll demonstrate with Claude: “claude-haiku-4-5-20251001”
