Implementation:Anthropics Anthropic sdk python Completion

Knowledge Sources	Anthropic SDK Python
Domains	API Types, Legacy Completions
Last Updated	2026-02-15 12:00 GMT

Overview

Completion is a Pydantic model that represents the response from the legacy Text Completions API. It contains the generated completion text, the model used, a unique identifier, stop reason, and a type discriminator.

Description

The Completion class extends BaseModel and models the response payload returned by the legacy /v1/complete endpoint. This is the older completion-style API (as opposed to the newer Messages API).

The model contains the following fields:

• id -- A unique string identifier for the completion. The format and length may change over time.
• completion -- The generated text, up to and excluding any stop sequences.
• model -- The model that produced the completion, typed as the Model type.
• stop_reason -- An optional string indicating why generation stopped. Values include "stop_sequence" (a stop sequence was reached) or "max_tokens" (the max_tokens_to_sample limit was exceeded).
• type -- A literal discriminator always set to "completion".

Usage

Use Completion when working with responses from the legacy Text Completions API. This model is returned by client.completions.create() when stream=False. Note that the Text Completions API is a legacy interface; the Messages API is recommended for new development.

Code Reference

Source Location

• Repository: Anthropic SDK Python
• File: src/anthropic/types/completion.py

Signature

class Completion(BaseModel):
    id: str
    completion: str
    model: Model
    stop_reason: Optional[str] = None
    type: Literal["completion"]

Import

from anthropic.types import Completion

I/O Contract

Fields

Field	Type	Required	Description
id	str	Yes	Unique object identifier.
completion	str	Yes	The generated completion text, excluding stop sequences.
model	Model	Yes	The model that produced the completion.
stop_reason	Optional[str]	No	Why generation stopped: "stop_sequence" or "max_tokens".
type	Literal["completion"]	Yes	Object type. Always "completion".

Usage Examples

import anthropic

client = anthropic.Anthropic()

# Legacy Text Completions API
completion = client.completions.create(
    model="claude-2.1",
    max_tokens_to_sample=256,
    prompt="\n\nHuman: What is the capital of France?\n\nAssistant:",
)

print(f"ID: {completion.id}")
print(f"Model: {completion.model}")
print(f"Completion: {completion.completion}")
print(f"Stop reason: {completion.stop_reason}")
print(f"Type: {completion.type}")

Related Pages

• CompletionCreateParams -- Request parameters for the legacy Text Completions API that produces this response.