> ## Documentation Index
> Fetch the complete documentation index at: https://docs.abliteration.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Anthropic Messages

> Tool calling on /v1/messages — input_schema definition, tool_use response blocks, tool_result continuation blocks.

`POST /v1/messages`

Tools are declared with a flat `input_schema` (no outer `type: "function"` wrapper). The model returns `tool_use` content blocks.

## Define a tool

```python theme={"system"}
tools = [{
    "name": "get_weather",
    "description": "Get the weather for a city",
    "input_schema": {
        "type": "object",
        "properties": {"city": {"type": "string"}},
        "required": ["city"],
    },
}]
```

## Full loop

```python theme={"system"}
from anthropic import Anthropic

client = Anthropic(base_url="https://api.abliteration.ai", auth_token=os.environ["ABLIT_KEY"])

messages = [{"role": "user", "content": "What's the weather in Lagos?"}]

resp = client.messages.create(
    model="abliterated-model",
    max_tokens=1024,
    tools=tools,
    messages=messages,
)

messages.append({"role": "assistant", "content": resp.content})

tool_results = []
for block in resp.content:
    if block.type == "tool_use":
        result = get_weather(block.input["city"])  # your function
        tool_results.append({
            "type": "tool_result",
            "tool_use_id": block.id,
            "content": str(result),
        })

messages.append({"role": "user", "content": tool_results})

final = client.messages.create(
    model="abliterated-model",
    max_tokens=1024,
    tools=tools,
    messages=messages,
)
print(final.content[0].text)
```

## Streaming

Stream with `stream=True`. Tool-use blocks arrive as `content_block_start` with `type: "tool_use"`, followed by `input_json_delta` events. `stop_reason: "tool_use"` signals the turn needs a follow-up with `tool_result`.

## Forcing a tool

```python theme={"system"}
tool_choice={"type": "tool", "name": "get_weather"}
```

`{type: "any"}` forces some tool; `{type: "auto"}` is the default.

## Parallel tool calls

Multiple `tool_use` blocks can appear in the same response. Return one `tool_result` block per `tool_use_id` in the next user turn.