​
Instructions

The Claude series models can be accessed via the official native API. Before using, make sure to install or upgrade the anthropic dependency: 
pip install -U anthropic
For non-Claude models, please use the OpenAI API format instead.

​
Models Information

ModelClaude Opus 4Claude Sonnet 4Claude Sonnet 3.7Claude Sonnet 3.5Claude Haiku 3.5Claude Opus 3Claude Haiku 3
Extended ThinkingYesYesYesNoNoNoNo
Context Window200K200K200K200K200K200K200K
Max Output32000 tokens64000 tokens64000 tokens8192 tokens8192 tokens4096 tokens4096 tokens
Training Cut-offMar 2025Mar 2025Nov 2024Apr 2024July 2024Aug 2023Aug 2023
  1. For models version 3.5 and above, if you need output longer than 4096 tokens, be sure to explicitly specify the "max_tokens" parameter, referring to the Max Output column in the table above.
  2. For Sonnet 3.7, you can increase the max output from 64K to 128K by passing extra_headers={"anthropic-beta": "output-128k-2025-02-19"}. See the “Streaming 128K” example below.

​
Claude 4 New Features

​
New Refusal Stop Reason

Claude 4 models introduce a new refusal stop reason for content that the model declines to generate for safety reasons: 
{
  "id": "msg_014XEDjypDjFzgKVWdFUXxZP",
  "type": "message",
  "role": "assistant",
  "model": "claude-sonnet-4-20250514",
  "content": [{"type": "text", "text": "I would be happy to assist you. You can "}],
  "stop_reason": "refusal",
  "stop_sequence": null,
  "usage": {
    "input_tokens": 564,
    "cache_creation_input_tokens": 0,
    "cache_read_input_tokens": 0,
    "output_tokens": 22
  }
}
When migrating to Claude 4, you should update your application to handle refusal stop reasons.

​
Extended Thinking

With extended thinking enabled, the Messages API for Claude 4 models returns a summary of Claude’s full thinking process. Summarized thinking provides the full intelligence benefits of extended thinking, while preventing misuse. While the API is consistent across Claude 3.7 and 4 models, streaming responses for extended thinking might return in a “chunky” delivery pattern, with possible delays between streaming events. Summarization is processed by a different model than the one you target in your requests. The thinking model does not see the summarized output.

​
Interleaved Thinking

Claude 4 models support interleaving tool use with extended thinking, allowing for more natural conversations where tool uses and responses can be mixed with regular messages. Interleaved thinking is in beta. To enable interleaved thinking, add the beta header interleaved-thinking-2025-05-14 to your API request: 
extra_headers={
    "anthropic-beta": "interleaved-thinking-2025-05-14"
}
Endpoint: POST /v1/messages

​
Usage

curl https://aihubmix.com/v1/messages \ # Replace the official endpoint with AiHubMix's API endpoint
     --header "x-api-key: $AIHUBMIX_API_KEY" \ # Replace with the key you generated in AiHubMix
     --header "anthropic-version: 2023-06-01" \
     --header "content-type: application/json" \
     --data \
'{
    "model": "claude-3-5-sonnet-20241022",
    "max_tokens": 1024,
    "messages": [
        {"role": "user", "content": "Hello, world"}
    ]
}'

​
Request Body

{
  "model": "claude-3-5-sonnet-20241022",
  "max_tokens": 1024,
  "messages": [
    {
      "role": "user",
      "content": "What is the meaning of life?"
    }
  ]
}

​
Request Parameters

NameLocationTypeRequiredDescription
x-api-keyheaderstringNoBearer AIHUBMIX_API_KEY
Content-TypeheaderstringNonone
bodybodyobjectNonone
» modelbodystringYesnone
» messagesbody[object]Yesnone
»» rolebodystringNonone
»» contentbodystringYesnone
» max_tokensbodynumberYesnone

​
Response Example

200 Response

{
  "id": "msg_013Uf6CwwyjSe35n3yVaPbLM",
  "type": "message",
  "role": "assistant",
  "model": "claude-3-5-sonnet-20241022",
  "content": [
    {
      "type": "text",
      "text": "That's one of humanity's most enduring and complex philosophical questions! While there's no universal answer, I aim to explore such questions thoughtfully while acknowledging their complexity. I try to focus on having meaningful conversations and helping where I can. What does meaning in life mean to you?"
    }
  ],
  "stop_reason": "end_turn",
  "stop_sequence": null,
  "usage": {
    "input_tokens": 14,
    "cache_creation_input_tokens": 0,
    "cache_read_input_tokens": 0,
    "output_tokens": 61
  }
}

​
Response Results

Status CodeStatus DescriptionDescriptionData Model
200OKnoneInline

​
Migrating to Claude 4

If you’re migrating from Claude 3.7 to Claude 4 models, please note the following changes:

​
Update Model Names

# From Claude 3.7
model="claude-3-7-sonnet-20250219"

# Migrate to Claude 4
model="claude-sonnet-4-20250514"  # or "claude-opus-4-20250514"

​
Handle New Stop Reasons

Update your application to handle the new refusal stop reason: 
if response.stop_reason == "refusal":
    print("Claude refused to generate this content")
elif response.stop_reason == "end_turn":
    print("Completed normally")

​
Remove Unsupported Features

  • Token-efficient tool use: Only available in Claude Sonnet 3.7, no longer supported in Claude 4
  • Extended output: The output-128k-2025-02-19 beta header is only available in Claude Sonnet 3.7
If you’re migrating from Claude Sonnet 3.7, we recommend removing these beta headers from your requests: 
# Remove these headers (if present)
# "token-efficient-tools-2025-02-19"
# "output-128k-2025-02-19"

​
Using Claude in Applications (Example: Lobe-Chat)

Here’s how you can configure Claude models in a third-party application like Lobe-Chat:
  1. Navigate to the settings page and select Claude as your model provider.
  2. Enter your API Key from AiHubMix.
  3. Set the API proxy endpoint to: 
    https://aihubmix.com
  4. (Recommended) Enable the “Client Request Mode” option.
  5. Add your chosen model to the model list.
    • It’s recommended to copy the model name from AiHubMix’s settings page and paste it in the application.
图片 图片