Anthropic Messages

POST /v1/messages mirrors the Anthropic Messages API. It is the right choice if your project is built on the Anthropic SDK, or if you run a tool that expects Anthropic's format, Claude Code being the main example. Only Claude models are available here. Every other chat model goes through /v1/chat/completions.

Claude models on this endpoint

Model id	Price
`claude-opus-5.5`	$3.78 / $18.90 per 1M tokens
`claude-sonnet-5`	$3.24 / $16.20 per 1M tokens
`claude-fable-5.1`	$9.45 / $47.25 per 1M tokens

Send any other model id and the request is refused with an error.

Headers and body

http

POST https://kadraris.com/v1/messages
x-api-key: sk-kd-...
anthropic-version: 2023-06-01
Content-Type: application/json

Every request must carry anthropic-version: 2023-06-01. If you use an official SDK you do not have to think about it, because the SDK adds it. Your key may be sent as x-api-key or as Authorization: Bearer (see Authentication).

Field	Type	Required	Meaning
`model`	string	yes	A Claude model id
`max_tokens`	integer	yes	Most tokens the model may write
`messages`	array	yes	`user` and `assistant` turns, alternating, Anthropic format
`system`	string or array	no	The system prompt
`stream`	boolean	no	`true` for Anthropic-style Server-Sent Events
`temperature`	number	no	Sampling temperature
`tools`	array	no	Tools the model may use, Anthropic format

Sending a message

bash

curl https://kadraris.com/v1/messages \
  -H "x-api-key: $KADRARIS_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-opus-5.5",
    "max_tokens": 1024,
    "system": "You review code briefly and point out bugs.",
    "messages": [{"role": "user", "content": "Review: def is_even(n): return n % 2 == 1"}]
  }'

python

# pip install anthropic
import os
from anthropic import Anthropic

client = Anthropic(base_url="https://kadraris.com", api_key=os.environ["KADRARIS_API_KEY"])

msg = client.messages.create(
    model="claude-opus-5.5",
    max_tokens=1024,
    system="You review code briefly and point out bugs.",
    messages=[{"role": "user", "content": "Review: def is_even(n): return n % 2 == 1"}],
)
print(msg.content[0].text)
print(msg.usage)

javascript

// npm install @anthropic-ai/sdk
import Anthropic from "@anthropic-ai/sdk";

const client = new Anthropic({
  baseURL: "https://kadraris.com",
  apiKey: process.env.KADRARIS_API_KEY,
});

const msg = await client.messages.create({
  model: "claude-opus-5.5",
  max_tokens: 1024,
  system: "You review code briefly and point out bugs.",
  messages: [{ role: "user", content: "Review: def is_even(n): return n % 2 == 1" }],
});
console.log(msg.content[0].text);

Watch the base URL. For the Anthropic SDK it is https://kadraris.com with no /v1 at the end, since the SDK adds /v1/messages on its own.

What the API returns

json

{
  "id": "msg_...",
  "type": "message",
  "role": "assistant",
  "model": "claude-opus-5.5",
  "content": [{ "type": "text", "text": "The check is inverted: n % 2 == 1 is true for odd numbers..." }],
  "stop_reason": "end_turn",
  "usage": { "input_tokens": 31, "output_tokens": 57 }
}

Streamed replies

Setting stream: true switches the reply to a sequence of Anthropic events, among them message_start, content_block_delta and message_stop.

bash

curl -N https://kadraris.com/v1/messages \
  -H "x-api-key: $KADRARIS_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{"model": "claude-sonnet-5", "max_tokens": 512, "stream": true, "messages": [{"role": "user", "content": "Describe a CDN in two sentences."}]}'

python

with client.messages.stream(
    model="claude-sonnet-5",
    max_tokens=512,
    messages=[{"role": "user", "content": "Describe a CDN in two sentences."}],
) as stream:
    for text in stream.text_stream:
        print(text, end="", flush=True)

javascript

const stream = client.messages.stream({
  model: "claude-sonnet-5",
  max_tokens: 512,
  messages: [{ role: "user", content: "Describe a CDN in two sentences." }],
});
stream.on("text", (text) => process.stdout.write(text));
await stream.finalMessage();

Tools

A tool is described with name, description and input_schema, as in Anthropic's API. When the model wants to use one, it returns a tool_use block. You run the tool and answer with a tool_result block in the next user turn.

python

tools = [{
    "name": "get_order_status",
    "description": "Look up where an order is in shipping",
    "input_schema": {
        "type": "object",
        "properties": {"order_id": {"type": "string"}},
        "required": ["order_id"],
    },
}]

messages = [{"role": "user", "content": "Has order B-2210 shipped yet?"}]
first = client.messages.create(model="claude-sonnet-5", max_tokens=1024, tools=tools, messages=messages)

if first.stop_reason == "tool_use":
    call = next(b for b in first.content if b.type == "tool_use")
    result = {"order_id": call.input["order_id"], "status": "shipped"}  # replace with a real lookup
    messages += [
        {"role": "assistant", "content": first.content},
        {"role": "user", "content": [{"type": "tool_result", "tool_use_id": call.id, "content": str(result)}]},
    ]
    final = client.messages.create(model="claude-sonnet-5", max_tokens=1024, tools=tools, messages=messages)
    print(final.content[0].text)

Error responses

Errors come back in Anthropic's format, and the Anthropic SDK turns them into its standard exceptions:

json

{ "type": "error", "error": { "type": "authentication_error", "message": "Invalid API key." } }

Status codes match the rest of the Kadraris API. By default one key may make 600 requests per minute. More in Errors.

Choosing between the two chat endpoints

Your client	Endpoint
Anthropic SDK, Claude Code	`/v1/messages`
OpenAI SDK, Cursor, Cline, Continue, n8n, Make	`/v1/chat/completions`
OpenAI SDK, but you want a Claude model	`/v1/chat/completions` with `model: "claude-..."`