release: 1.1.0

[stainless-app]

Automated Release PR

1.1.0 (2026-03-04)

Full Changelog: v1.0.0...v1.1.0

Features

• api: add uploads (multipart) subresource to files (f510cc9)
• api: api update (e95f5f1)
• api: api update (eb27c60)
• api: api update (c4be5ae)
• api: api update (153c4ba)
• api: api update (216baf6)
• api: api update (8fef561)
• api: api update (10a541d)
• api: files uploads (multipart) subresource (93a98e6)
• api: manual updates (e38a505)
• api: manual updates (860868e)
• api: manual updates (33abdc9)
• use multipart uploads API for large files automatically with optional config (#18) (ff0f83f)

Bug Fixes

• cancel orphaned upload tasks and fix file size calc for io bytes (#21) (46aea64)
• count uploaded bytes correctly even if parts finish out of order (#20) (eb8af86)
• do not block event loop for read_part on async coroutine in multipart upload (#19) (537f825)

Chores

• format all api.md files (d167e2d)
• internal: add request options to SSE classes (f241bf6)
• internal: fix lint error on Python 3.14 (72f3d05)
• internal: make test_proxy_environment_variables more resilient (d0ec2d9)
• internal: make test_proxy_environment_variables more resilient to env (c287c24)
• update mock server docs (b3523db)

---

This pull request is managed by Stainless's GitHub App.

The semver version number is based on included commit messages. Alternatively, you can manually set the version number in the title of this pull request.

For a better experience, it is recommended to use either rebase-merge or squash-merge when merging this pull request.

🔗 Stainless website
📚 Read the docs
🙋 Reach out for help or questions

---

Note

Medium Risk
Introduces new multipart upload flow and changes default behavior of files.create to conditionally use it for large files, which could affect upload correctness/performance and error handling. Additional model/schema updates are low risk but broad and may impact downstream typing/serialization expectations.

Overview
Adds multipart file uploads. Introduces a new files.uploads subresource (create, retrieve, list, abort, complete) with generated types/tests and docs updates.

Changes files.create behavior. client.files.create() (sync/async) now optionally/automatically uses multipart uploads for large files (default threshold 100MB) via new MultipartUploadOptions/PartUploadEvent, including concurrent part uploads, progress callbacks, and best-effort abort on failure.

Misc API/internal updates. Propagates request options into SSE stream classes, tweaks proxy env var tests for resilience, expands several response models (e.g., image metadata/external IDs, parsing job image field), updates search option schemas/docs, and bumps version metadata to 1.1.0.

^{Written by Cursor Bugbot for commit 53d0bc7. This will update automatically on new commits. Configure here.}

[stainless-app]

🧪 Testing

To try out this version of the SDK:

pip install 'https://pkg.stainless.com/s/mixedbread-python/46aea647e9aa19ee8095dd1bafe8c54d9f1d7bb9/mixedbread-1.0.0-py3-none-any.whl'

Expires at: Fri, 03 Apr 2026 19:08:41 GMT
Updated at: Wed, 04 Mar 2026 19:08:41 GMT

[cursor]

Unquoted find expansion breaks doc formatting

Low Severity

format:docs uses $(find . -type f -name api.md) unquoted inside bash -c, so any api.md path containing whitespace or shell-special characters gets split into multiple arguments, causing scripts/utils/ruffen-docs.py to receive invalid filenames.

 

[cursor]

Find command searches through excluded directories

Low Severity

The find . -type f -name api.md command recursively searches all directories including .venv, .git, __pycache__, and .mypy_cache. This causes unnecessary performance overhead searching through large dependency directories and could format unintended files if they happen to be named api.md. The command should exclude these directories using -not -path patterns or limit search depth with -maxdepth.

 

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

[cursor]

Multipart upload silently drops user request options

Medium Severity

When the multipart upload path is taken (either via auto-detection with multipart_upload=None or explicitly), the extra_headers, extra_query, extra_body, and timeout parameters are silently ignored. The call to multipart_create_sync/multipart_create_async only forwards the uploads resource, file, and _opts — none of the user-provided request options. This is especially surprising in the default auto-detect case, where a user consistently passing timeout=600 would see it respected for small files but silently dropped for files exceeding the 100MB threshold.

Additional Locations (1)

• src/mixedbread/resources/files/files.py#L440-L456