Skill בנייה עם OpenAI API ל-Claude

מתי להשתמש

"OpenAI API", "GPT-4 API", "Whisper", "DALL-E API", "Assistants API".

הוראות עבודה

1. Setup

from openai import OpenAI

client = OpenAI(api_key="sk-...")

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Hello"}]
)
print(response.choices[0].message.content)

2. Models 2026

3. Function Calling

tools = [{
    "type": "function",
    "function": {
        "name": "get_weather",
        "description": "Get weather",
        "parameters": {
            "type": "object",
            "properties": {"location": {"type": "string"}}
        }
    }
}]

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Weather in Tel Aviv?"}],
    tools=tools
)

4. Structured Outputs (JSON Mode)

Old Way

response_format={"type": "json_object"}

New (better) — Strict JSON Schema

from pydantic import BaseModel

class Person(BaseModel):
    name: str
    age: int
    email: str

response = client.beta.chat.completions.parse(
    model="gpt-4o",
    messages=[...],
    response_format=Person
)
person = response.choices[0].message.parsed

5. Vision

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{
        "role": "user",
        "content": [
            {"type": "text", "text": "What's in this image?"},
            {"type": "image_url", "image_url": {"url": "https://..."}}
        ]
    }]
)

6. Whisper (Audio → Text)

audio_file = open("interview.mp3", "rb")
transcript = client.audio.transcriptions.create(
    model="whisper-1",
    file=audio_file
)
print(transcript.text)

Supports

• 50+ languages including Hebrew.
• $0.006 per minute.

7. Text-to-Speech (TTS)

response = client.audio.speech.create(
    model="tts-1",
    voice="alloy",  # 6 voices
    input="Hello, this is AI speaking"
)
response.stream_to_file("output.mp3")

8. DALL-E (Images)

response = client.images.generate(
    model="dall-e-3",
    prompt="A serene Tel Aviv sunset",
    size="1024x1024",
    quality="standard"
)
url = response.data[0].url

9. Embeddings

response = client.embeddings.create(
    input="Your text here",
    model="text-embedding-3-small"  # or large
)
embedding = response.data[0].embedding  # 1536-dim vector

Use For

• Semantic search.
• RAG.
• Clustering.

10. Assistants API (Beta)

What

• Persistent assistants with memory + tools.
• Code Interpreter, Retrieval, Function calling.
• Threads (persistent conversations).

When

• Building chatbot with memory.
• Need built-in code execution.
• Want OpenAI-managed RAG.

Caveat

• Less flexible than building yourself.
• Vendor lock-in.

11. Fine-tuning

# Upload training data (JSONL)
file = client.files.create(
    file=open("training.jsonl", "rb"),
    purpose="fine-tune"
)

# Create job
job = client.fine_tuning.jobs.create(
    training_file=file.id,
    model="gpt-4o-mini-2024-07-18"
)

When

• Domain-specific tone.
• High-volume specific task.
• 100+ training examples.

12. Best Practices

• Use gpt-4o-mini for 80% of tasks.
• Stream for UX.
• Structured outputs — reliable parsing.
• Retry logic for rate limits.
• API key in env, not code.

13. Cost Optimization

• Cheaper models first.
• Cache common responses (your side).
• Batch API for async (50% off).
• Limit max_tokens.

14. OpenAI vs Anthropic

15. אסיים בהמלצה.

פרומפט לדוגמה

Build voice transcription pipeline with Whisper.

Function calling for sales agent. Code.

Structured outputs vs JSON mode — when each.

© 2026 AI Expert Pro | גרסה 1.0.0