15 AI API Use Cases with Complete Code Examples

AI APIs power an incredible range of applications — from simple chatbots to complex document analysis pipelines. This guide covers 15 real-world use cases with production-ready code examples, cost estimates, and model recommendations for each scenario. All examples use Token1.US for unified access to GPT-4o, Claude, DeepSeek-V3, and GLM-4-Plus.

1. Customer Service Chatbot

The most common AI API use case. Route customer questions to the cheapest model that handles them well — typically DeepSeek-V3 at $0.14/$0.28 per million tokens.

def customer_service_bot(user_message, context=""):
    response = client.chat.completions.create(
        model="deepseek-v3",  # Cost-efficient for Q&A
        messages=[
            {"role": "system", "content": "You are a helpful customer service agent..."},
            {"role": "user", "content": f"Context: {context}\n\nQuestion: {user_message}"}
        ],
        max_tokens=300
    )
    return response.choices[0].message.content

Cost: ~$0.001 per conversation (500 input + 200 output tokens on DeepSeek-V3)

2. Content Generation (Blog Posts, Articles)

Generate long-form content with Claude 3.5 Sonnet for the best writing quality, or DeepSeek-V3 for cost efficiency.

def generate_article(topic, word_count=1000, style="professional"):
    response = client.chat.completions.create(
        model="claude-3-5-sonnet",  # Best writing quality
        messages=[
            {"role": "system", "content": f"Write a {style} article about {topic}..."},
            {"role": "user", "content": f"Write approximately {word_count} words."}
        ],
        max_tokens=2000,
        temperature=0.7  # Creative but coherent
    )
    return response.choices[0].message.content

Cost: ~$0.03 per article on Claude, ~$0.001 on DeepSeek-V3

3. Code Assistant & Auto-Completion

Build a programming assistant that explains code, suggests fixes, and generates functions. Claude 3.5 Sonnet is the top choice for coding tasks.

def code_assistant(code_snippet, task="explain"):
    response = client.chat.completions.create(
        model="claude-3-5-sonnet",
        messages=[
            {"role": "system", "content": "You are an expert programmer. " + task},
            {"role": "user", "content": f"```\n{code_snippet}\n```"}
        ],
        max_tokens=1000
    )
    return response.choices[0].message.content

4. Text Summarization

Condense long documents into concise summaries. Use GLM-4-Plus for a balance of quality and cost.

def summarize(text, max_words=100):
    response = client.chat.completions.create(
        model="glm-4-plus",
        messages=[
            {"role": "system", "content": f"Summarize in {max_words} words or less."},
            {"role": "user", "content": text}
        ],
        max_tokens=200
    )
    return response.choices[0].message.content

5. Sentiment Analysis

Classify text sentiment at scale. DeepSeek-V3 is perfect for this — fast, cheap, and accurate for classification tasks.

def analyze_sentiment(text):
    response = client.chat.completions.create(
        model="deepseek-v3",
        messages=[
            {"role": "system", "content": "Respond with only: positive, negative, or neutral"},
            {"role": "user", "content": text}
        ],
        max_tokens=10,
        temperature=0  # Deterministic
    )
    return response.choices[0].message.content.strip().lower()

Cost: ~$0.0001 per analysis — classify 1 million texts for ~$100

6. Translation & Localization

Translate content across languages with GPT-4o for highest accuracy or GLM-4-Plus for Chinese/English tasks.

def translate(text, target_language="Chinese"):
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[
            {"role": "system", "content": f"Translate to {target_language}. Output only the translation."},
            {"role": "user", "content": text}
        ],
        max_tokens=1000
    )
    return response.choices[0].message.content

7. Retrieval-Augmented Generation (RAG)

Combine your knowledge base with AI for accurate, sourced answers. Use Claude's 200K context window for large document sets.

def rag_query(question, relevant_docs):
    context = "\n\n".join(relevant_docs)
    response = client.chat.completions.create(
        model="claude-3-5-sonnet",  # 200K context window
        messages=[
            {"role": "system", "content": "Answer based ONLY on the provided context. If the answer isn't in the context, say so."},
            {"role": "user", "content": f"Context:\n{context}\n\nQuestion: {question}"}
        ],
        max_tokens=500
    )
    return response.choices[0].message.content

8. Email Drafting & Reply

Automate email composition and responses. DeepSeek-V3 handles routine emails at minimal cost.

def draft_email_reply(incoming_email, tone="professional"):
    response = client.chat.completions.create(
        model="deepseek-v3",
        messages=[
            {"role": "system", "content": f"Write a {tone} email reply."},
            {"role": "user", "content": f"Original email:\n{incoming_email}\n\nDraft a reply."}
        ],
        max_tokens=300
    )
    return response.choices[0].message.content

9. Data Extraction & Structuring

Extract structured data (JSON) from unstructured text using function calling.

def extract_entities(text):
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[{"role": "user", "content": f"Extract entities as JSON: {text}"}],
        functions=[{
            "name": "save_entities",
            "parameters": {
                "type": "object",
                "properties": {
                    "names": {"type": "array", "items": {"type": "string"}},
                    "dates": {"type": "array", "items": {"type": "string"}},
                    "amounts": {"type": "array", "items": {"type": "number"}}
                }
            }
        }],
        function_call={"name": "save_entities"}
    )
    return response.choices[0].message.function_call.arguments

10. Product Description Generator

Generate SEO-optimized e-commerce product descriptions from basic specs.

def generate_product_description(product_name, features, keywords):
    response = client.chat.completions.create(
        model="glm-4-plus",  # Good balance of quality and cost
        messages=[{
            "role": "user",
            "content": f"Product: {product_name}\nFeatures: {features}\nSEO keywords: {keywords}\n\nWrite a compelling 150-word product description."
        }],
        max_tokens=250
    )
    return response.choices[0].message.content

11. Social Media Content

Generate posts for Twitter/X, LinkedIn, Instagram with platform-specific formatting.

def social_post(topic, platform="twitter"):
    limits = {"twitter": 280, "linkedin": 3000, "instagram": 2200}
    response = client.chat.completions.create(
        model="deepseek-v3",
        messages=[{
            "role": "user",
            "content": f"Write a {platform} post about {topic}. Max {limits.get(platform, 500)} chars."
        }],
        max_tokens=200
    )
    return response.choices[0].message.content

12. Document Q&A System

Upload PDFs or documents and let users ask questions. Combine with vector search for large document sets.

def document_qa(document_text, question):
    response = client.chat.completions.create(
        model="gpt-4o",  # 128K context for large docs
        messages=[
            {"role": "system", "content": "Answer questions based on this document."},
            {"role": "user", "content": f"Document:\n{document_text[:100000]}\n\nQ: {question}"}
        ],
        max_tokens=500
    )
    return response.choices[0].message.content

13. Intent Classification for NLP

Classify user intents for routing, analytics, or automation workflows.

def classify_intent(user_input):
    response = client.chat.completions.create(
        model="deepseek-v3",
        messages=[
            {"role": "system", "content": "Classify intent: purchase, support, info, complaint, other. Respond with one word."},
            {"role": "user", "content": user_input}
        ],
        max_tokens=10,
        temperature=0
    )
    return response.choices[0].message.content.strip().lower()

14. Image Analysis (Multimodal)

Use GPT-4o's vision capabilities to analyze images, extract text, or describe visual content.

def analyze_image(image_base64, question="Describe this image"):
    response = client.chat.completions.create(
        model="gpt-4o",  # Only model with vision support
        messages=[{
            "role": "user",
            "content": [
                {"type": "text", "text": question},
                {"type": "image_url", "image_url": {"url": f"data:image/png;base64,{image_base64}"}}
            ]
        }],
        max_tokens=300
    )
    return response.choices[0].message.content

15. Content Moderation

Automatically flag inappropriate, spam, or harmful content using AI classification.

def moderate_content(text):
    response = client.chat.completions.create(
        model="deepseek-v3",
        messages=[
            {"role": "system", "content": "Classify as: safe, spam, harassment, violence, sexual, hate. Respond with JSON: {category, confidence, reason}"},
            {"role": "user", "content": text}
        ],
        max_tokens=100,
        temperature=0
    )
    return response.choices[0].message.content

Cost Optimization Tips for Production

Model Routing Strategy

Don't use GPT-4o for everything. Implement smart routing: use DeepSeek-V3 for 70% of requests (simple tasks), GLM-4-Plus for 20%, and GPT-4o/Claude only for the 10% that truly need premium quality. This can reduce costs by 80-90%.

Caching

Cache identical queries. If 1,000 users ask "What is Token1.US?", serve the cached response instead of making 1,000 API calls. Redis or in-memory caching works well.

Batch Processing

For bulk tasks, use batch mode instead of streaming. See our streaming vs batch guide for cost comparisons.

Token Optimization

Shorten prompts, remove unnecessary context, and set appropriate max_tokens limits. Every token costs money. Read our token pricing guide for optimization strategies.

Implementation Resources

Ready to build? Get your API key and start with free trial credits.