3.2.1

What's Changed

• Allow empty choices list by @vrslev in #26

Full Changelog: 3.2.0...3.2.1

3.2.0

What's Changed

• Handle too long context errors for multimodal vLLM models by @vrslev in #25

Full Changelog: 3.1.1...3.2.0

3.1.1

What's Changed

• Fix raising validation errors when streaming by @vrslev in #24

Full Changelog: 3.1.0...3.1.1

3.1.0

What's Changed

• Reraise LLMResponseValidationError on pydantic validation errors by @vrslev in #23

Full Changelog: 3.0.0...3.1.0

3.0.0

What's Changed

• Add reasoning content by @insani7y in #22

New Contributors

• @insani7y made their first contribution in #22

Full Changelog: 2.3.0...3.0.0

2.3.0

What's Changed

• Add image support for OpenAI client by @vrslev in #21

Full Changelog: 2.2.0...2.3.0

2.2.0

What's Changed

• Allow to set default temperature in LLM model config & request extra in YandexGPT config by @vrslev in #19
• Use community-workflow by @vrslev in #17

Full Changelog: 2.1.0...2.2.0

2.1.0

What's Changed

• Let request params be changed via model config by @mrkaaa in #16
• Update setup-uv to v5 by @vrslev in #15

New Contributors

• @mrkaaa made their first contribution in #16

Full Changelog: 2.0.0...2.1.0

2.0.0

What's Changed

• Move back to HTTPX by @vrslev in #13
• Stream chunks, not partial responses by @vrslev in #12

Full Changelog: 1.3.0...2.0.0

1.3.0

What's Changed

The most significant change is moving to Niquests. It means that generally requests will be faster. It also means that extra keyword arguments passed to any_llm_client.get_client() have to be rewritten for Niquests.

• Remove noqa comments from examples by @vrslev in #9
• Update ruff by @vrslev in #10
• Use niquests by @vrslev in #7

Full Changelog: 1.2.0...1.3.0