Fix: Claude API Error 400 (2026)

The 400 invalid_request_error is the most common Claude API error. It means something in your request format or content is wrong. This guide covers every known cause and the fix for each.

The Error

{
 "type": "error",
 "error": {
 "type": "invalid_request_error",
 "message": "There was an issue with the format or content of your request."
 },
 "request_id": "req_011CSHoEeqs5C35K2UUqR7Fy"
}

Quick Fix

  1. Check that model, max_tokens, and messages are all present in your request.
  2. Verify messages alternate between user and assistant roles.
  3. If using Claude Opus 4.6, remove any prefilled assistant messages (prefill is not supported).

What Causes This

The 400 error covers all request validation failures. The API checks your request format before processing. Common triggers include:

  • Missing required parameters (model, max_tokens, messages).
  • Invalid message role sequence (messages must alternate user/assistant).
  • Prefilling assistant messages on models that do not support it.
  • Invalid tool definitions or malformed JSON schemas.
  • Exceeding the 100,000 message limit per request.
  • Using tool_choice values incompatible with extended thinking (only auto or none work).

Full Solution

Validate Required Parameters

Every Messages API request needs these three fields:

For more on this topic, see Fix: Claude Can’t Open This Chat Error.

import anthropic
client = anthropic.Anthropic()
message = client.messages.create(
 model="claude-sonnet-4-6", # Required: valid model ID
 max_tokens=1024, # Required: positive integer
 messages=[ # Required: non-empty array
 {"role": "user", "content": "Hello, Claude"}
 ]
)

Fix Message Role Alternation

Messages must alternate between user and assistant. Two consecutive user messages will trigger a 400 error:

# WRONG: Two user messages in a row
messages = [
 {"role": "user", "content": "Hello"},
 {"role": "user", "content": "Are you there?"} # Error!
]
# CORRECT: Alternating roles
messages = [
 {"role": "user", "content": "Hello"},
 {"role": "assistant", "content": "Hi there!"},
 {"role": "user", "content": "Are you there?"}
]

Fix Prefill Not Supported Error

Claude Opus 4.6 does not support prefilling assistant messages. If you send a request with a prefilled last assistant message, you get a 400 invalid_request_error:

# WRONG on Opus 4.6: prefilled assistant message
messages = [
 {"role": "user", "content": "Write JSON"},
 {"role": "assistant", "content": "{"} # Error on Opus 4.6!
]
# CORRECT: Use output_config instead
message = client.messages.create(
 model="claude-opus-4-6",
 max_tokens=1024,
 messages=[{"role": "user", "content": "Write JSON for a user profile"}],
 output_config={
 "format": {
 "type": "json_schema",
 "schema": {
 "type": "object",
 "properties": {
 "name": {"type": "string"},
 "email": {"type": "string"}
 },
 "required": ["name", "email"],
 "additionalProperties": False
 }
 }
 }
)

Handle Extended Thinking + Tool Use

When using extended thinking with tools, only tool_choice: auto or tool_choice: none is supported:

# WRONG: tool_choice "any" with thinking
message = client.messages.create(
 model="claude-sonnet-4-6",
 max_tokens=16000,
 thinking={"type": "enabled", "budget_tokens": 10000},
 tool_choice={"type": "any"}, # Error!
 tools=[...],
 messages=[...]
)
# CORRECT: Use "auto" with thinking
message = client.messages.create(
 model="claude-sonnet-4-6",
 max_tokens=16000,
 thinking={"type": "enabled", "budget_tokens": 10000},
 tool_choice={"type": "auto"}, # OK
 tools=[...],
 messages=[...]
)

Catch the Error in Code

import anthropic
client = anthropic.Anthropic()
try:
 message = client.messages.create(
 model="claude-sonnet-4-6",
 max_tokens=1024,
 messages=[{"role": "user", "content": "Hello"}]
 )
except anthropic.BadRequestError as e:
 print(f"400 Error: {e.message}")
 print(f"Request ID: {e.response.headers.get('request-id')}")

import Anthropic from "@anthropic-ai/sdk";
const client = new Anthropic();
try {
 await client.messages.create({
 model: "claude-sonnet-4-6",
 max_tokens: 1024,
 messages: [{ role: "user", content: "Hello" }]
 });
} catch (err) {
 if (err instanceof Anthropic.BadRequestError) {
 console.log(`400 Error: ${err.message}`);
 }
}

Prevention

  1. Use the SDK: The Python and TypeScript SDKs validate request structure before sending, catching many 400 errors client-side.
  2. Read the error message: The API message field in the error response explains exactly which parameter is invalid.
  3. Include the request_id: Every error response includes a request_id. Include it when contacting Anthropic support.
  4. Check model compatibility: Not all parameters work with all models. Extended thinking requires budget_tokens >= 1024 and budget_tokens < max_tokens.
I hit this exact error six months ago. Then I wrote a CLAUDE.md that tells Claude my stack, my conventions, and my error handling patterns. Haven't seen it since. I run 5 Claude Max subs, 16 Chrome extensions serving 50K users, and bill $500K+ on Upwork. These CLAUDE.md templates are what I actually use. Not theory — production configs. **[Grab the templates — $99 once, free forever →](https://zovo.one/lifetime?utm_source=ccg&utm_medium=cta-error&utm_campaign=claude-api-error-400-invalidrequesterror-explained)** 47/500 founding spots. Price goes up when they're gone.

Which model? → Take the 5-question quiz in our Model Selector.

Try it: Paste your error into our Error Diagnostic for an instant fix.