Ollama: REST API Fundamentals

curl http://localhost:11434/api/chat -d '{
  "model": "qwen2.5",
  "messages": [{"role": "user", "content": "Refund for order #12345"}]
}'

# Non-streaming generate request
curl http://localhost:11434/api/generate -d '{
  "model": "llama3.2",
  "prompt": "Write a haiku about coding",
  "stream": false
}'

# Response (abbreviated)
# {
#   "model": "llama3.2",
#   "response": "Lines of logic flow,\nBug hides in the deep syntax—\nSemicolon found.",
#   "done": true,
#   "total_duration": 2500000000,
#   "eval_count": 18
# }

{"status":"ok","data":{}}

# Chat with message history
curl http://localhost:11434/api/chat -d '{
  "model": "qwen2.5",
  "messages": [
    {"role": "system", "content": "You are a helpful customer service agent."},
    {"role": "user", "content": "I want to return my order #12345"},
    {"role": "assistant", "content": "I can help with that. May I ask the reason for the return?"},
    {"role": "user", "content": "The product arrived damaged"}
  ],
  "stream": false
}'

{"status":"ok","data":{}}

{"model":"llama3.2","message":{"role":"assistant","content":"I"},"done":false}
{"model":"llama3.2","message":{"role":"assistant","content":" can"},"done":false}
{"model":"llama3.2","message":{"role":"assistant","content":" help"},"done":false}
{"model":"llama3.2","message":{"role":"assistant","content":""},"done":true,"total_duration":1500000000}

# Streaming request with real-time output
curl http://localhost:11434/api/chat -d '{
  "model": "llama3.2",
  "messages": [{"role": "user", "content": "Hello!"}],
  "stream": true
}' | while read -r line; do
    # Extract content field from each NDJSON line
    echo "$line" | python3 -c "
import sys, json
data = json.load(sys.stdin)
if data.get('message', {}).get('content'):
    print(data['message']['content'], end='', flush=True)
"
done

{"status":"ok","data":{}}

# Low temperature: deterministic output
curl http://localhost:11434/api/generate -d '{
  "model": "llama3.2",
  "prompt": "What is 2+2?",
  "stream": false,
  "options": {"temperature": 0.1}
}'
# Response: "2+2 equals 4."

# High temperature: creative output
curl http://localhost:11434/api/generate -d '{
  "model": "llama3.2",
  "prompt": "What is 2+2?",
  "stream": false,
  "options": {"temperature": 1.5}
}'
# Response: "In the realm of mathematics, 2+2 opens the door to 4..."

{"status":"ok","data":{}}

# Force JSON output format
curl http://localhost:11434/api/chat -d '{
  "model": "llama3.2",
  "messages": [
    {"role": "system", "content": "You are a product catalog API. Return JSON only."},
    {"role": "user", "content": "List 3 laptops under $500"}
  ],
  "format": "json",
  "stream": false,
  "options": {"temperature": 0.3}
}'

# Response is valid JSON
# {"products":[{"name":"Acer Aspire 5","price":449,"spec":"8GB RAM, 256GB SSD"},...]}

{"status":"ok","data":{}}

#!/bin/bash
# ============================================
# Comprehensive: SupportBot API prototype
# Multi-turn customer service via REST API
# ============================================

API="http://localhost:11434/api/chat"
MODEL="qwen2.5"

# Function: Send a chat message and extract response
chat() {
    local system_prompt="$1"
    local user_msg="$2"
    local temp="${3:-0.4}"

    curl -s "$API" -d "$(cat <<EOF
{
  "model": "$MODEL",
  "messages": [
    {"role": "system", "content": "$system_prompt"},
    {"role": "user", "content": "$user_msg"}
  ],
  "stream": false,
  "options": {"temperature": $temp, "num_ctx": 4096}
}
EOF
)" | python3 -c "import sys,json; print(json.load(sys.stdin)['message']['content'])"
}

# Customer service system prompt
SYSTEM="You are SupportBot, a customer service agent for an e-commerce store. Be polite, concise, and helpful. If you cannot answer, say 'Let me connect you with a human agent.'"

# Simulate customer interactions
echo "=== Query 1: Order Status ==="
chat "$SYSTEM" "Where is my order #88765? It has been 5 days."

echo ""
echo "=== Query 2: Return Request ==="
chat "$SYSTEM" "I received a damaged item. Order #12345. I want a refund."

echo ""
echo "=== Query 3: Product Question ==="
chat "$SYSTEM" "Does the wireless headphone support Bluetooth 5.3?"

echo ""
echo "=== Benchmark ==="
time chat "$SYSTEM" "Hello" > /dev/null

=== Query 1: Order Status ===
I'd be happy to check on your order #88765. Based on our tracking system, your order is currently in transit and expected to arrive within 2-3 business days. You can track it at track.example.com/88765.

=== Query 2: Return Request ===
I'm sorry to hear about the damaged item. For order #12345, I've initiated a return request. You'll receive a prepaid shipping label via email within 24 hours. Once we receive the item, a full refund will be processed within 3-5 business days.

=== Query 3: Product Question ===
Yes, our wireless headphones support Bluetooth 5.3 with a range of up to 15 meters. They also feature active noise cancellation and 30-hour battery life.