Claude 4 Migration Cheat Sheet
Everything you need to switch in under 60 seconds. Bookmark this page.
TL;DR — What to switch to
If you use Claude 4 Opus
→ Claude Opus 4.8
Save 67% ($15→$5 input, $75→$25 output)
5x context window (200K→1M). Same provider, same API key.
If you use Claude Sonnet 4
→ Claude Sonnet 4.6
Same price, 5x context
$3/$15 per 1M tokens. Context: 200K→1M. Strict upgrade.
Cheapest cross-provider option
→ DeepSeek V4 Pro
Save 97% ($0.435/$0.87 per 1M tokens)
1M context. Requires new API key from DeepSeek.
Best quality-to-price ratio
→ GPT-5
Save 88% input, 87% output
$1.25/$10 per 1M tokens. 272K context. Requires OpenAI key.
The 1-Line Code Change
If you're staying with Anthropic, your API key and SDK don't change. Just update the model string:
// Python — Anthropic SDK
client = anthropic.Anthropic()
# Before (retired June 15):
response = client.messages.create(
model="claude-4-opus-20250514",
...
)
# After (recommended):
response = client.messages.create(
model="claude-opus-4-8-20260115",
...
)
client = anthropic.Anthropic()
# Before (retired June 15):
response = client.messages.create(
model="claude-4-opus-20250514",
...
)
# After (recommended):
response = client.messages.create(
model="claude-opus-4-8-20260115",
...
)
// JavaScript — Anthropic SDK
// Before:
model: "claude-4-opus-20250514"
// After:
model: "claude-opus-4-8-20260115"
// Before:
model: "claude-4-opus-20250514"
// After:
model: "claude-opus-4-8-20260115"
# If switching to DeepSeek (cheapest option):
# Install: pip install openai
client = openai.OpenAI(
api_key="your-deepseek-key",
base_url="https://api.deepseek.com/v1"
)
response = client.chat.completions.create(
model="deepseek-v4-pro",
...
)
# Install: pip install openai
client = openai.OpenAI(
api_key="your-deepseek-key",
base_url="https://api.deepseek.com/v1"
)
response = client.chat.completions.create(
model="deepseek-v4-pro",
...
)
Cost Comparison (per 1M tokens)
| Model | Provider | Input | Output | Context | vs Claude 4 Opus |
|---|---|---|---|---|---|
| Claude 4 Opus | Anthropic | $15.00 | $75.00 | 200K | — |
| Claude Sonnet 4 | Anthropic | $3.00 | $15.00 | 200K | — |
| Claude Opus 4.8 | Anthropic | $5.00 | $25.00 | 1M | -67% |
| Claude Sonnet 4.6 | Anthropic | $3.00 | $15.00 | 1M | -80% input |
| Claude Haiku 4.5 | Anthropic | $1.00 | $5.00 | 200K | -93% |
| GPT-5 | OpenAI | $1.25 | $10.00 | 272K | -87% |
| GPT-5 mini | OpenAI | $0.25 | $2.00 | 272K | -97% |
| Gemini 2.5 Pro | $1.25 | $10.00 | 1M | -87% | |
| DeepSeek V4 Pro | DeepSeek | $0.435 | $0.87 | 1M | -97% |
| DeepSeek V4 Flash | DeepSeek | $0.14 | $0.28 | 1M | -99% |
| Mistral Large 3 | Mistral | $0.50 | $1.50 | 262K | -97% |
| Llama 4 Maverick | Meta (Together.ai) | $0.27 | $0.85 | 1M | -98% |
Which replacement should you pick?
"I want to stay with Anthropic"
Use Claude Opus 4.8 (for Opus users) or Claude Sonnet 4.6 (for Sonnet users). Same API key, same SDK. Just change the model name. Opus 4.8 is 67% cheaper with 5x context.
"I want the absolute cheapest option"
Use DeepSeek V4 Pro ($0.435/$0.87). 97% cheaper than Claude 4 Opus. 1M context window. Requires a new API key from DeepSeek. Supports OpenAI SDK format.
"I want the best quality-to-price ratio"
Use GPT-5 ($1.25/$10) or Gemini 2.5 Pro ($1.25/$10). Both are 87% cheaper than Claude 4 Opus with strong reasoning capabilities.
"I need to migrate fast with minimal code changes"
Use Claude Sonnet 4.6. It's the same price as Sonnet 4, same API key, same SDK. Just change one string. Zero risk migration.
"I use Claude for code generation / agentic tasks"
Use Claude Opus 4.8 or GPT-5. Both excel at complex reasoning and code generation. Opus 4.8 has 1M context for large codebases.
Migration Checklist
- Identify all places in your code that reference claude-4-opus or claude-sonnet-4
- Choose your replacement model (see decision tree above)
- Update the model name string in your code
- If switching providers: get a new API key and update base_url
- Test with a small request to verify the new model works
- Run your full test suite against the new model
- Update any hardcoded token limits (new models may have different defaults)
- Monitor costs for the first few days to verify expected savings
- Update your documentation and team runbooks
- Remove deprecated model references from config files
Status: Claude 4 models were retired June 15, 2026. API calls to Claude 4 models now return errors.
Calculate your exact savings
Use the APIpulse migration tool to enter your current Claude 4 usage and see exactly how much you'll save with each alternative — down to the penny.
Free Migration Calculator Free Audit — Save More →All users get optimization recommendations that save up to 40% more. 100% free.
Want this as an email checklist?
Get the complete migration checklist emailed to you — with code snippets, cost comparisons, and a deadline tracker.
Get the Free Checklist →More Claude 4 migration resources:
🔴 Claude 4 Has Been Shut Down — Read the Latest: