feat(mistral): add support for OpenAI-style response format dictionaries with Mistral

[HareeshBahuleyan]

Current support is limited to params.response_format being a Pydantic Model. This PR enables using openAI schema. Will be useful when migrating openai agents framework to use any-llm: mozilla-ai/any-agent#828

[codecov]

Codecov Report

❌ Patch coverage is 83.33333% with 1 line in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/any_llm/providers/mistral/mistral.py	83.33%	0 Missing and 1 partial ⚠️

Files with missing lines	Coverage Δ
src/any_llm/providers/mistral/mistral.py	82.05% <83.33%> (-13.90%)	⬇️

... and 33 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[HareeshBahuleyan]

Ran CI tests:
https://github.com/mozilla-ai/any-llm/actions/runs/18411357927/job/52464763632
The failures observed there I guess are unrelated - they seem to be for other providers, not Mistral, which is being changed in this PR.

[daavoo]

Thanks @HareeshBahuleyan !

Could you add a test to https://github.com/mozilla-ai/any-llm/blob/main/tests/unit/providers/test_mistral_provider.py ? You can see in other providers examples of mocking the internal client

[HareeshBahuleyan]

@daavoo A unit test has been added now 👍

[daavoo]

@HareeshBahuleyan I think we should not be mocking this mistral functions, we should actually try to see if they work with the inputs we are passing.

Then, we can assert on the arguments passed to mocked_mistralchat.complete_async

Suggested change

	patch("any_llm.providers.mistral.mistral.response_format_from_pydantic_model") as mocked_pydantic_converter,
	patch("any_llm.providers.mistral.mistral.ResponseFormat") as mocked_response_format,

[HareeshBahuleyan]

Sure, updated the code to actually call the functions and verify that they pass the response_format argument with the expected schema to the completion() function.

[daavoo]

🚀