qlawmarq

主要AI/LLMモデル比較表（2025年8月時点）

モデル	コンテキストウィンドウ	出力トークン	価格（百万トークンあたり）	最適化・得意分野	情報源
OpenAI GPT-5	272K（入力） 400K（合計）	128K	入力: $1.25 出力: $10	統合推論システム、コーディング、マルチモーダル	openai.com
GPT-5 mini	272K（入力）	128K	入力: $0.25 出力: $2	リアルタイム体験、推論	openai.com
GPT-5 nano	272K（入力）	128K	入力: $0.05 出力: $0.40	軽量タスク、最高コスト効率	openai.com
GPT-4o	128K	16.4K	入力: $2.50 出力: $10	マルチモーダル、リアルタイム音声・画像	openai.com
GPT-4o mini	128K	16.4K	入力: $0.15 出力: $0.60	高速・軽量タスク、コスト効率	openai.com
GPT-4	8K/32K	4K	入力: $30出力: $60	汎用的推論、レガシー	[openai.com](https://platform.openai.co

	#####
	#
	# GitHub Actions will deploy your app to Cloud Run in Google Cloud Platform (GCP) when you push to master or main branch.
	#
	# 1. Set up your GCP project and Cloud Run service account
	# 2. Ensure your service account has permissions to push images to Artifact Registry and deploy to Cloud Run.
	# 3. Create a key.json file for the service account and add it to your GitHub repo as a secret. (In this example, the secret name is GCP_DEPLOYER_KEY)
	# 4. Enable the Cloud Run API and Artifact Registry API in your GCP project
	# 5. Add this file as .github/workflows/deployer.yml
	# 6. Add a Dockerfile to your repo to build your app

	# The following is an example of a command to clone an development(ex-app-development) as an production(ex-app-production).
	kubectl get deployment ex-app-development -o json \
	\| jq '.metadata.name = "ex-app-production"' \
	\| kubectl apply -f -

	from langchain.llms import OpenAI
	from langchain import OpenAI, SQLDatabase, SQLDatabaseChain
	from langchain.prompts.prompt import PromptTemplate
	import os

	def chat_with_mysql_database(prompt: str):
	"""
	Args:
	prompt (str): Question to ask e.g. "Generate an SQL query to retrieve all users"
	Returns:

	import random
	import string

	def generate_random_string():
	selects = random.sample(string.ascii_uppercase, 3) + random.sample(
	string.digits, 3) + random.sample(string.ascii_lowercase, 3)
	random.shuffle(selects)
	unique_code = ''.join(selects)
	print(unique_code)
	return unique_code

-name: Remove old versions of Google App Engine
-#####
-#
-# Add this file to .github/workflows/delete-old-gae.yml then GitHub Actions will remove old versions of Google App Engine.
-#
-#####
-on:
-  workflow_dispatch:

	name: Deployer

	#####
	#
	# Add this file to .github/workflows/google_app_engine_deployer.yml, then GitHub Actions will detect any updates on master/main branch and deploy changes.
	# This example uses Node.js 18. You can choose any language you like.
	# The build command is also `npm run build', change it as you like.
	#
	# 1. Set up service account for deployment and create key.json and set it to env variable
	# 2. Create your app.yaml if you never deploy the app to GAE