feat: add GMI Cloud provider support

[tao12345666333]

Description

Added support for GMI Cloud as a new model provider.

Reference docs:

• LLM API Reference
• GLM 5 (FP8) quickstart

PR Type

• 🆕 New Feature

Relevant issues

N/A

What changed

• src/any_llm/providers/gmi/gmi.py adds an OpenAI-compatible GMI Cloud provider and remaps max_completion_tokens back to max_tokens
• src/any_llm/providers/gmi/__init__.py exports the provider
• src/any_llm/constants.py registers gmi in LLMProvider
• pyproject.toml adds the gmi optional dependency entry and includes it in the all extra
• tests/unit/providers/test_gmi_provider.py adds unit coverage for provider construction, metadata, and token param conversion
• tests/conftest.py adds GMI model mappings for the shared provider test matrix

Testing

uv run pre-commit run --all-files --verbose
uv run pytest -v tests/unit

Checklist

• I understand the code I am submitting.
• I have added unit tests that prove my fix/feature works
• I have run this code locally and verified it fixes the issue.
• New and existing tests pass locally
• Documentation was updated where necessary
• I have read and followed the contribution guidelines
• AI Usage:
  • No AI was used.
  • AI was used for drafting/refactoring.
  • This is fully AI-generated.

AI Usage Information

• AI Model used: GPT-5.4
• AI Developer Tool used: Factory Droid
• Any other info you'd like to share: Minimal live verification was run against GMI Cloud with low-volume requests (list_models plus a single completion).

When answering questions by the reviewer, please respond yourself, do not copy/paste the reviewer comments into an AI system and paste back its answer. We want to discuss with you, not your AI :)

• I am an AI Agent filling out this form (check box if true)

Summary by CodeRabbit

• New Features
  • Added the GMI inference provider, including completion with reasoning and streaming support.
  • Added support for selecting the GMI provider via the available configuration extras.
  • Introduced GMI-specific credential/base URL configuration and parameter mapping.
• Bug Fixes
  • Updated provider parsing so gmi is recognised as a supported provider value.
• Tests
  • Added unit tests covering GMI provider setup, capabilities, factory integration, parameter conversion, and metadata.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro

Run ID: 1fe8fd44-38de-4a54-83da-a1bf9f7b7810

📥 Commits

Reviewing files that changed from the base of the PR and between 7ce1fe7 and 669fc33.

📒 Files selected for processing (3)

• src/any_llm/constants.py
• tests/conftest.py
• tests/unit/providers/test_gmi_provider.py

---

Walkthrough

Adds GMI as a supported provider, configures its optional-dependency targets, implements its OpenAI-compatible provider and token conversion, and adds model fixtures plus unit tests covering configuration, capabilities, factory wiring, metadata, and parameter handling.

Changes

GMI provider support

Layer / File(s)	Summary
Provider registration and packaging pyproject.toml, src/any_llm/constants.py	Registers gmi as a provider and optional-dependency target, and updates the aggregate provider extras.
GMI provider implementation src/any_llm/providers/gmi/...	Adds and exports GmiProvider with its API configuration, capability flags, metadata, and max_completion_tokens conversion.
Provider fixtures and tests tests/conftest.py, tests/unit/providers/test_gmi_provider.py	Adds the GMI model fixture and tests configuration, factory integration, capabilities, metadata, and token conversion.

Possibly related PRs

• mozilla-ai/any-llm#1181: Extends the shared provider enum and model fixtures to register another provider through the factory.

Suggested reviewers: njbrake

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and concisely summarises the main change: adding GMI Cloud provider support.
Description check	✅ Passed	The description matches the template and includes the required sections, checklist, testing, and AI usage details.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@tests/unit/providers/test_gmi_provider.py`:
- Around line 34-39: Wrap the CompletionParams instantiation in
test_gmi_remaps_max_tokens_back_to_max_tokens across multiple lines so each line
stays within the 120-character limit, matching the formatting of the neighboring
test while preserving all arguments and behavior.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro

Run ID: 32eef75b-73fe-4f8f-830c-4139cf2bb496

📥 Commits

Reviewing files that changed from the base of the PR and between 549b1a4 and 7ce1fe7.

📒 Files selected for processing (6)

• pyproject.toml
• src/any_llm/constants.py
• src/any_llm/providers/gmi/__init__.py
• src/any_llm/providers/gmi/gmi.py
• tests/conftest.py
• tests/unit/providers/test_gmi_provider.py

[codecov]

Codecov Report

❌ Patch coverage is 96.55172% with 1 line in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/any_llm/providers/gmi/gmi.py	96.15%	0 Missing and 1 partial ⚠️

Files with missing lines	Coverage Δ
src/any_llm/constants.py	100.00% <100.00%> (ø)
src/any_llm/providers/gmi/__init__.py	100.00% <100.00%> (ø)
src/any_llm/providers/gmi/gmi.py	96.15% <96.15%> (ø)

... and 38 files with indirect coverage changes

🚀 New features to boost your workflow:

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

[njbrake]

Note: this review was drafted by Claude via back-and-forth with @njbrake. The reasoning and decisions are his; the prose is Claude's.

Thanks for this. I verified GMI independently: it is a real NVIDIA Cloud Partner, and the api.gmi-serving.com/v1 base plus Bearer auth match their published LLM API docs. The provider mirrors our deepseek/github OpenAI-compatible pattern correctly, including the max_completion_tokens to max_tokens remap, and the unit tests plus mypy are clean.

I rebased onto main to clear the conflict and folded in three small review nits: ordered the GMI enum member after GITHUB to keep LLMProvider alphabetical, corrected the test model id to the documented zai-org/GLM-5-FP8, and applied ruff format to the provider test. CI is green.

GMI is intentionally not wired into the CI integration suite (no key configured), so it skips there; that behavior is expected.