LLM & Agentic Model Comparison

A guide to the different AI models you can use with your chat app, including pricing, API key setup, and code examples.

Quick Comparison Table

Provider Details

1. OpenAI (GPT-4o, GPT-4o-mini)

Best for: Default choice. Great quality, easy API, huge community.

Getting an API key: 1. Go to https://platform.openai.com/api-keys 2. Click "Create new secret key" 3. Copy the key (starts with sk-...) 4. Add to .env:

OPENAI_API_KEY=sk-...

Free credit: New accounts get $5-50 free credit (varies).

Code example:

from openai import OpenAI
import os

client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))

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

Pricing example: 1,000 chat conversations × 500 input tokens + 500 output tokens = ($0.15 × 0.5M + $0.60 × 0.5M) = $0.075 + $0.30 = $0.38

2. Anthropic (Claude 3.5 Sonnet, Claude 3 Haiku)

Best for: Complex reasoning tasks, long documents, safety-conscious apps.

Getting an API key: 1. Go to https://console.anthropic.com/ 2. Sign up and go to API Keys 3. Create a key (starts with sk-ant-...) 4. Add to .env:

ANTHROPIC_API_KEY=sk-ant-...

Free credit: New accounts get $5-10 free credit.

Code example:

from anthropic import Anthropic
import os

client = Anthropic(api_key=os.getenv("ANTHROPIC_API_KEY"))

response = client.messages.create(
    model="claude-3-haiku-20240307",  # or "claude-3-5-sonnet-20241022"
    max_tokens=1000,
    messages=[
        {"role": "user", "content": "Hello!"}
    ]
)
print(response.content[0].text)

3. Google (Gemini 2.0 Flash)

Best for: Cheapest option, good quality, handles images + text.

Getting an API key: 1. Go to https://aistudio.google.com/apikey 2. Click "Create API key" 3. Copy the key 4. Add to .env:

GOOGLE_API_KEY=AIza...

Free tier: Very generous — 1,500 requests/day free.

Code example:

import google.generativeai as genai
import os

genai.configure(api_key=os.getenv("GOOGLE_API_KEY"))

model = genai.GenerativeModel("gemini-2.0-flash")
response = model.generate_content("Hello!")
print(response.text)

4. DeepSeek

Best for: Strong reasoning at a low price. Open-weight model (you can download it).

Getting an API key: 1. Go to https://platform.deepseek.com/ 2. Sign up and go to API Keys 3. Create a key 4. Add to .env:

DEEPSEEK_API_KEY=sk-...

Code example:

from openai import OpenAI  # DeepSeek uses OpenAI-compatible API
import os

client = OpenAI(
    api_key=os.getenv("DEEPSEEK_API_KEY"),
    base_url="https://api.deepseek.com"
)

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

5. Meta Llama 3 (via Groq)

Best for: Open source, extremely fast inference (Groq has custom hardware).

Getting an API key: 1. Go to https://console.groq.com/keys 2. Create a free API key 3. Add to .env:

GROQ_API_KEY=gsk_...

Free tier: Very generous — 30 requests/minute, 14,400/day free.

Code example:

from openai import OpenAI  # Groq uses OpenAI-compatible API
import os

client = OpenAI(
    api_key=os.getenv("GROQ_API_KEY"),
    base_url="https://api.groq.com/openai/v1"
)

response = client.chat.completions.create(
    model="llama-3.1-70b-versatile",
    messages=[
        {"role": "user", "content": "Hello!"}
    ]
)
print(response.choices[0].message.content)

6. Mistral

Best for: EU data sovereignty requirements, multilingual apps.

Getting an API key: 1. Go to https://console.mistral.ai/ 2. Sign up and go to API Keys 3. Create a key 4. Add to .env:

MISTRAL_API_KEY=...

Code example:

from mistralai import Mistral
import os

client = Mistral(api_key=os.getenv("MISTRAL_API_KEY"))

response = client.chat.complete(
    model="mistral-large-latest",
    messages=[
        {"role": "user", "content": "Hello!"}
    ]
)
print(response.choices[0].message.content)

How to Swap Models in Your Chat App

In backend/app/routes.py, the chat endpoint calls an LLM. To swap providers:

Current code (OpenAI):

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=messages
)

Swap to Anthropic:

# Change the import at the top
from anthropic import Anthropic
client = Anthropic(api_key=os.getenv("ANTHROPIC_API_KEY"))

# Change the API call
response = client.messages.create(
    model="claude-3-haiku-20240307",
    max_tokens=1000,
    messages=messages
)
reply = response.content[0].text

Swap to DeepSeek:

# Keep OpenAI import, change client setup
client = OpenAI(
    api_key=os.getenv("DEEPSEEK_API_KEY"),
    base_url="https://api.deepseek.com"
)

# Same call structure
response = client.chat.completions.create(
    model="deepseek-chat",
    messages=messages
)
reply = response.choices[0].message.content

Cost Estimation Guide

Formula: (Avg input tokens + Avg output tokens) × Total conversations × Token cost

Example: If each chat uses 500 input + 300 output tokens, and you use GPT-4o-mini: - Per chat: (0.0005M × $0.15) + (0.0003M × $0.60) = $0.000075 + $0.00018 = $0.000255 - 10,000 chats: $2.55