Claude Messages API — TokSpan Docs

Endpoint

http
POST https://api.tokspan.com/v1/messages

Quick Examples

Use the Anthropic SDK with TokSpan as the base URL — no code changes beyond configuration:

from anthropic import Anthropic

client = Anthropic(
    api_key="sk-your-key",
    base_url="https://api.tokspan.com",  # ← host root; the SDK appends /v1/messages
)

response = client.messages.create(
    model="MODEL_NAME",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello!"}],
)

print(response.content[0].text)

shell
curl -X POST "https://api.tokspan.com/v1/messages" \
  -H "x-api-key: sk-your-key" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "MODEL_NAME",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

json — Response
{
  "id": "msg_abc123",
  "type": "message",
  "role": "assistant",
  "content": [{
    "type": "text",
    "text": "Hello! How can I help you today?"
  }],
  "model": "MODEL_NAME",
  "stop_reason": "end_turn",
  "usage": {
    "input_tokens": 10,
    "output_tokens": 12
  }
}

Request Body

Parameter	Type	Required	Description
model	string	Yes	Claude model ID (e.g., `claude-opus-4-8`, `claude-sonnet-4-6`).
max_tokens	integer	Yes	Maximum tokens to generate. <strong>Required</strong> for the Messages API.
messages	array	Yes	Array of messages alternating between <code>user</code> and <code>assistant</code> roles. The first message must be from the user.
system	string	No	System prompt — a string or array of content blocks.
temperature	number	No	Sampling temperature (0–1).
stream	boolean	No	Enable SSE streaming. Default: <code>false</code>.
top_p	number	No	Nucleus sampling (0–1).
tools	array	No	Tool definitions for tool use. See Tool Calling.
tool_choice	object	No	Control tool selection: `{"type": "auto"}` or a specific tool.
thinking	object	No	Extended thinking config: `{"type": "enabled", "budget_tokens": 1024}`. Budget must be ≥1024 and less than `max_tokens`.

System Prompts

Pass the system prompt as a top-level <code>system</code> field (not a message):

json
{
  "model": "MODEL_NAME",
  "max_tokens": 1024,
  "system": "You are a helpful assistant that always answers in French.",
  "messages": [{"role": "user", "content": "What is the capital of France?"}]
}

Streaming (SSE)

Set stream: true to receive Anthropic SSE events (message_start, content_block_delta, message_stop, etc.).

shell
curl -X POST "https://api.tokspan.com/v1/messages" \
  -H "x-api-key: sk-your-key" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "MODEL_NAME",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "Tell me a story."}],
    "stream": true
  }'

Configuration: Set the Anthropic SDK's base_url to https://api.tokspan.com — the SDK appends /v1/messages itself. Authentication accepts both x-api-key + anthropic-version headers (Anthropic-native) and Authorization: Bearer. The Anthropic SDK handles this automatically.