Implementation:Mistralai Client python Files Upload

Knowledge Sources	Mistral Client Python Mistral AI Docs
Domains	Fine_Tuning, File_Management
Last Updated	2026-02-15 14:00 GMT

Overview

Concrete tool for uploading training data files to Mistral AI servers provided by the Files resource.

Description

The Files.upload() and Files.upload_async() methods upload a file to Mistral's servers using multipart form encoding. The file is wrapped in a File object (with file_name and content as bytes). An optional purpose parameter categorizes the file (e.g., "fine-tune"). The method returns an UploadFileOut object with the file's id (needed for job creation), filename, bytes count, and created_at timestamp.

Usage

Call client.files.upload() with a File object containing the JSONL training data. Use the returned id in TrainingFile objects for fine_tuning.jobs.create().

Code Reference

Source Location

Repository: client-python
File: src/mistralai/client/files.py
Lines: L22-122 (sync), L124-224 (async)

Signature

class Files:
    def upload(
        self,
        *,
        file: Union[File, FileTypedDict],
        purpose: Optional[FilePurpose] = None,
    ) -> UploadFileOut:
        ...

    async def upload_async(
        self,
        *,
        file: Union[File, FileTypedDict],
        purpose: Optional[FilePurpose] = None,
    ) -> UploadFileOut:
        ...

Import

from mistralai import Mistral
from mistralai.models import File
# Access via: client.files.upload(...)

I/O Contract

Inputs

Name	Type	Required	Description
file	File	Yes	File object with file_name and content (bytes)
purpose	Optional[FilePurpose]	No	File purpose (e.g., "fine-tune")

Outputs

Name	Type	Description
result	UploadFileOut	Contains id, filename, bytes, created_at
result.id	str	File ID for use in fine-tuning jobs

Usage Examples

Upload Training File

from mistralai import Mistral
from mistralai.models import File

client = Mistral(api_key="your-key")

# Read and upload training data
with open("train.jsonl", "rb") as f:
    uploaded = client.files.upload(
        file=File(file_name="train.jsonl", content=f.read()),
        purpose="fine-tune",
    )

print(f"File ID: {uploaded.id}")
print(f"Size: {uploaded.bytes} bytes")

Related Pages

Implements Principle

Principle:Mistralai_Client_python_File_Upload

Requires Environment

Environment:Mistralai_Client_python_Python_SDK_Environment

Uses Heuristic

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment