Claude 4 Shutdown Checklist: Complete Developer Guide

This is the only guide you need. Whether you're preparing for the Claude 4 shutdown or already seeing 410 errors, this checklist covers everything: diagnosis, migration, cost optimization, and verification. Most developers complete this in 5-15 minutes.

Timeline: What's Happening

PRE-SHUTDOWN CHECKLIST (Do Before June 15)

Complete these steps before the shutdown to avoid downtime:

# Find all Claude 4 references in your codebase
grep -r "claude-4-opus\|claude-4-sonnet\|claude-4" \
  --include="*.js" --include="*.ts" --include="*.py" \
  --include="*.json" --include="*.env" --include="*.yaml" \
  --include="*.yml" --include="*.toml" --include="*.cfg" .

# Also check for model IDs in config files
grep -r "model.*claude-4\|claude-4.*model" --include="*.json" --include="*.env" .

# Python
# Before (returns 410 Gone)
client.messages.create(model="claude-4-opus", ...)

# After (official successor, 67% cheaper)
client.messages.create(model="claude-opus-4-8", ...)

# Node.js
// Before
{ model: "claude-4-opus" }
// After
{ model: "claude-opus-4-8" }

# Environment variable
ANTHROPIC_MODEL=claude-opus-4-8

# Python
# Before (returns 410 Gone)
client.messages.create(model="claude-4-sonnet", ...)

# After (official successor, 50% cheaper)
client.messages.create(model="claude-sonnet-4-6", ...)

# Node.js
// Before
{ model: "claude-4-sonnet" }
// After
{ model: "claude-sonnet-4-6" }

# Environment variable
ANTHROPIC_MODEL=claude-sonnet-4-6

# Quick test (Python)
import anthropic
client = anthropic.Anthropic()  # Uses ANTHROPIC_API_KEY env var

response = client.messages.create(
    model="claude-opus-4-8",  # or claude-sonnet-4-6
    max_tokens=100,
    messages=[{"role": "user", "content": "Say 'migration successful'"}]
)
print(response.content[0].text)
# Should output: "migration successful"

# Quick test (Node.js)
const response = await client.messages.create({
    model: "claude-sonnet-4-6",
    max_tokens: 100,
    messages: [{ role: "user", content: "Say 'migration successful'" }]
});
console.log(response.content[0].text);
# Should output: "migration successful"

DURING SHUTDOWN: If You're Seeing 410 Errors

If your app is already broken, follow these steps immediately:

# Python error
anthropic.NotFoundError: Error code: 410 -
{'type': 'error', 'error': {'type': 'not_found_error',
'message': 'claude-4-opus has been retired. Please upgrade to
claude-opus-4-8. See https://docs.anthropic.com/docs/deprecations'}}

# Node.js error
Error: 410 model_not_found
"claude-4-sonnet is no longer available. Use claude-sonnet-4-6."

# curl response
HTTP/1.1 410 Gone
{"type":"error","error":{"type":"not_found_error","message":"claude-4-opus has been retired"}}

The Cost Breakdown: What You Were Paying vs What You'll Pay

Here's the silver lining of the Claude 4 shutdown: every replacement is cheaper.

10 Migration Guides: Choose Your Path

We've created detailed migration guides for every major alternative. Each includes step-by-step code, cost comparisons, and FAQPage schema for rich snippets:

Want to see all 42 alternatives with live pricing? Check our Claude 4 Migration Hub — it has 18 head-to-head comparisons, migration calculators, and code snippets for every alternative.

Choose Your Migration Path

Not sure which alternative is right for you? Here's a quick decision guide:

• ✓Want zero code changes? → Use claude-opus-4-8 or claude-sonnet-4-6 (official Anthropic successors, same API)
• ✓Want maximum savings? → Use DeepSeek V4 Flash ($0.14/$0.28 — 99% cheaper)
• ✓Want best quality-per-dollar? → Use DeepSeek V4 Pro ($0.44/$0.87 — near-Opus quality at 97% less)
• ✓Need largest context window? → Use Gemini 3.1 Pro (1M tokens) or Llama 4 Maverick (1M tokens)
• ✓Need EU/GDPR compliance? → Use Mistral Medium 3.5 (EU data sovereignty)
• ✓Want OpenAI ecosystem? → Use GPT-5 or GPT-5.5 Pro (largest AI ecosystem)
• ✓Want open source? → Use Llama 4 Maverick (fully open, self-hostable)

Migration Code Snippets

Here are ready-to-use code snippets for the most popular alternatives:

Option A: Stick with Anthropic (easiest)

# Python — Just change the model name
import anthropic
client = anthropic.Anthropic()

response = client.messages.create(
    model="claude-opus-4-8",  # was claude-4-opus
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello"}]
)

# Node.js
const response = await client.messages.create({
    model: "claude-sonnet-4-6",  # was claude-4-sonnet
    max_tokens: 1024,
    messages: [{ role: "user", content: "Hello" }]
});

Option B: Switch to DeepSeek (cheapest)

# Python — pip install openai
from openai import OpenAI
client = OpenAI(
    api_key="your-deepseek-key",
    base_url="https://api.deepseek.com/v1"
)

response = client.chat.completions.create(
    model="deepseek-v4-pro",  # or deepseek-v4-flash for even cheaper
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello"}]
)
print(response.choices[0].message.content)

# Node.js
import OpenAI from 'openai';
const client = new OpenAI({
    apiKey: 'your-deepseek-key',
    baseURL: 'https://api.deepseek.com/v1'
});

const response = await client.chat.completions.create({
    model: 'deepseek-v4-pro',
    max_tokens: 1024,
    messages: [{ role: 'user', content: 'Hello' }]
});

Option C: Switch to GPT-5 (OpenAI ecosystem)

# Python — pip install openai
from openai import OpenAI
client = OpenAI()  # Uses OPENAI_API_KEY env var

response = client.chat.completions.create(
    model="gpt-5",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello"}]
)
print(response.choices[0].message.content)

# Node.js
import OpenAI from 'openai';
const client = new OpenAI();

const response = await client.chat.completions.create({
    model: 'gpt-5',
    max_tokens: 1024,
    messages: [{ role: 'user', content: 'Hello' }]
});

Option D: Switch to Gemini (1M context)

# Python — pip install google-genai
from google import genai
client = genai.Client()  # Uses GEMINI_API_KEY env var

response = client.models.generate_content(
    model="gemini-3.1-pro",
    contents="Hello"
)
print(response.text)

# Node.js
import { GoogleGenAI } from '@google/genai';
const client = new GoogleGenAI({ apiKey: process.env.GEMINI_API_KEY });

const response = await client.models.generateContent({
    model: 'gemini-3.1-pro',
    contents: 'Hello'
});
console.log(response.text);

POST-SHUTDOWN VERIFICATION

After deploying your fix, verify everything works:

Key Questions on Shutdown Day