Implementation:Cohere ai Cohere python ApiMeta Model

Knowledge Sources	Cohere Python SDK
Domains	SDK, API Metadata
Last Updated	2026-02-15 14:00 GMT

Overview

The ApiMeta class is a Pydantic model representing metadata returned with Cohere API responses.

Description

The ApiMeta model provides structured access to metadata that accompanies Cohere API responses. It includes the API version information via ApiMetaApiVersion, a breakdown of billed units via ApiMetaBilledUnits, token usage statistics via ApiMetaTokens, a count of cached tokens that hit the inference cache, and any warnings generated during request processing. This model extends UncheckedBaseModel and is auto-generated from the Cohere API definition by Fern.

Usage

Use this model to inspect billing, token usage, API version, and warning information after making any Cohere API call. It is typically accessible as a meta attribute on API response objects.

Code Reference

Source Location

• Repository: Cohere Python SDK
• File: src/cohere/types/api_meta.py

Signature

class ApiMeta(UncheckedBaseModel):
    api_version: typing.Optional[ApiMetaApiVersion] = None
    billed_units: typing.Optional[ApiMetaBilledUnits] = None
    tokens: typing.Optional[ApiMetaTokens] = None
    cached_tokens: typing.Optional[float] = pydantic.Field(default=None)
    warnings: typing.Optional[typing.List[str]] = None

Import

from cohere.types import ApiMeta

I/O Contract

Field	Type	Required	Default	Description
api_version	Optional[ApiMetaApiVersion]	No	None	The API version information for the response.
billed_units	Optional[ApiMetaBilledUnits]	No	None	Breakdown of billed units by category (tokens, images, search units, classifications).
tokens	Optional[ApiMetaTokens]	No	None	Token usage statistics for the request.
cached_tokens	Optional[float]	No	None	The number of prompt tokens that hit the inference cache.
warnings	Optional[List[str]]	No	None	List of warning messages generated during request processing.

Usage Examples

import cohere

client = cohere.Client(api_key="YOUR_API_KEY")

response = client.chat(
    model="command-r-plus",
    message="Explain quantum computing in simple terms.",
)

# Access the ApiMeta object
meta = response.meta

if meta is not None:
    # Check API version
    if meta.api_version is not None:
        print(meta.api_version)

    # Inspect billed units
    if meta.billed_units is not None:
        print(f"Input tokens billed: {meta.billed_units.input_tokens}")
        print(f"Output tokens billed: {meta.billed_units.output_tokens}")

    # Check cached tokens
    if meta.cached_tokens is not None:
        print(f"Cached tokens: {meta.cached_tokens}")

    # Review any warnings
    if meta.warnings:
        for warning in meta.warnings:
            print(f"Warning: {warning}")

Related Pages

• Environment:Cohere_ai_Cohere_python_Python_SDK_Runtime