Anthropic API

Neosantara Anthropic Compatible

To meet the demand for using the Anthropic API ecosystem, our API has added support for the Anthropic API format. With simple configuration, you can integrate the capabilities of Neosantara Model into the Anthropic API ecosystem.

Use Neosantara in Claude Code

1. Install Claude Code

   npm i -g @anthropic-ai/claude-code
   
2. Config Environment Variables

   export ANTHROPIC_BASE_URL=https://api.neosantara.xyz/anthropic
   export ANTHROPIC_AUTH_TOKEN=${YOUR_API_KEY}
   export API_TIMEOUT_MS=600000
   export ANTHROPIC_MODEL=nusantara-base
   export ANTHROPIC_SMALL_FAST_MODEL=nusantara-base
   export CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC=1
   

   The API_TIMEOUT_MS parameter is configured to prevent excessively long outputs that could cause the Claude Code client to time out. Here, we set the timeout duration to 10 minutes.
3. Run the Claude Code

   cd my-project
   claude
   

Invoke Neosantara Model via Anthropic API

1. Install Anthropic SDK

   pip install anthropic
   
2. Config Environment Variables

   export ANTHROPIC_BASE_URL=https://api.neosantara.xyz/anthropic
   export ANTHROPIC_API_KEY=${NAI_API_KEY}
   
3. Invoke the API

   import anthropic
   
   client = anthropic.Anthropic()
   
   message = client.messages.create(
       model="deepseek-r1-llama-70b",
       max_tokens=1000,
       system="You are a helpful assistant.",
       messages=[
           {
               "role": "user",
               "content": [
                   {
                       "type": "text",
                       "text": "Hi, how are you?"
                   }
               ]
           }
       ]
   )
   print(message.content)
   

Compatibility Matrix

Parameter	Required	Type	Claude SDK	Default	Notes
model	✅ Yes	string	✅ Yes	-	Use Neosantara model instead
messages	✅ Yes	array	✅ Yes	-	Alternating user/assistant
max_tokens	✅ Yes	integer	✅ Yes	-	Min 1 token
temperature	❌ No	number	✅ Yes	1.0	Range: 0.0-1.0
stream	❌ No	boolean	✅ Yes	false	SSE format
top_p	❌ No	number	✅ Yes	1.0	Range: 0.0-1.0
top_k	❌ No	number	✅ Yes	-	Must be > 0
stop_sequences	❌ No	array	✅ Yes	[]	Max 5 sequences
system	❌ No	string/array	✅ Yes	-	Separate parameter
tools	❌ No	array	✅ Yes	[]	Function definitions
tool_choice	❌ No	string	✅ Yes	”auto”	auto/any/specific
metadata	❌ No	object	✅ Yes	Custom tracking
service_tier	❌ No	string	✅ Yes	”auto”	auto/standard_only
thinking	❌ No	object	✅ Yes	-	Extended thinking mode