feat: add support for gemini-3-pro-preview model

[claudiunicolaa]

User description

• Add gemini/gemini-3-pro-preview with 1,048,576 max tokens
• Add vertex_ai/gemini-3-pro-preview with 1,048,576 max tokens
• Add test coverage for both model variants
• Update documentation with usage examples for both variants

This enables users to utilize Google's Gemini 3 Pro Preview model through both Google AI Studio and Vertex AI providers with full 1M+ token context window support.

Resources:

• https://ai.google.dev/gemini-api/docs/gemini-3

---

PR Type

Enhancement

---

Description

• Add gemini-3-pro-preview model support for both Google AI and Vertex AI

• Configure 1,048,576 max tokens for both model variants

• Add parameterized test coverage for both provider implementations

• Update documentation with usage examples for Google AI variant

---

Diagram Walkthrough

flowchart LR
  A["gemini-3-pro-preview<br/>Model Support"] --> B["Model Registry<br/>Configuration"]
  A --> C["Test Coverage<br/>Parameterized Tests"]
  A --> D["Documentation<br/>Usage Examples"]
  B --> E["Google AI Studio<br/>1M+ tokens"]
  B --> F["Vertex AI<br/>1M+ tokens"]

Loading

File Walkthrough

	Relevant files
Configuration changes	__init__.py Register gemini-3-pro-preview models with max tokens          pr_agent/algo/init.py Add vertex_ai/gemini-3-pro-preview with 1,048,576 max tokens Add gemini/gemini-3-pro-preview with 1,048,576 max tokens Both entries placed in appropriate alphabetical positions in model registry +2/-0
Tests	test_get_max_tokens.py Add parameterized test for gemini-3-pro-preview variants  tests/unittest/test_get_max_tokens.py Add parameterized test for both gemini-3-pro-preview variants Test validates 1,048,576 max tokens for Google AI and Vertex AI providers Uses pytest parametrization to consolidate test cases +14/-0
Documentation	qodo_merge_models.md Document gemini-3-pro-preview Google AI variant usage        docs/docs/usage-guide/qodo_merge_models.md Add gemini/gemini-3-pro-preview to supported models list Add configuration example for Google AI Studio variant Maintains consistency with existing Vertex AI variant documentation +8/-0

---

[qodo-free-for-open-source-projects]

PR Compliance Guide 🔍

Below is a summary of compliance checks for this PR:

Security Compliance
🟢	No security concerns identified No security vulnerabilities detected by AI analysis. Human verification advised for critical code.
Ticket Compliance
⚪	🎫 No ticket provided Create ticket/issue
Codebase Duplication Compliance
⚪	Codebase context is not defined Follow the guide to enable codebase context checks.
Custom Compliance
🟢	Consistent Naming Conventions Objective: All new variables, functions, and classes must follow the project's established naming standards Status: Passed
	No Dead or Commented-Out Code Objective: Keep the codebase clean by ensuring all submitted code is active and necessary Status: Passed
	Robust Error Handling Objective: Ensure potential errors and edge cases are anticipated and handled gracefully throughout the code Status: Passed
	Single Responsibility for Functions Objective: Each function should have a single, well-defined responsibility Status: Passed
	When relevant, utilize early return Objective: In a code snippet containing multiple logic conditions (such as 'if-else'), prefer an early return on edge cases than deep nesting Status: Passed
Update

Compliance status legend

🟢 - Fully Compliant
🟡 - Partial Compliant
🔴 - Not Compliant
⚪ - Requires Further Human Verification
🏷️ - Compliance label

[qodo-free-for-open-source-projects]

PR Code Suggestions ✨

Explore these optional code suggestions:

Category	Suggestion	Impact
General	✅ Parametrize test for both variants Suggestion Impact: The test was converted to a parametrized pytest test over the two model strings, removing the duplicated assertions and using a single assertion for both variants. code diff: + @pytest.mark.parametrize("model", [ + "gemini/gemini-3-pro-preview", + "vertex_ai/gemini-3-pro-preview", + ]) + def test_gemini_3_pro_preview(self, monkeypatch, model): + fake_settings = type("", (), { + "config": type("", (), { + "custom_model_max_tokens": 0, + "max_model_tokens": 0, })() })() + monkeypatch.setattr(utils, "get_settings", lambda: fake_settings) + assert get_max_tokens(model) == 1048576 - monkeypatch.setattr(utils, "get_settings", lambda: fake_settings) - - # Test Google AI Studio variant - model_gemini = "gemini/gemini-3-pro-preview" - expected = 1048576 - assert get_max_tokens(model_gemini) == expected - - # Test Vertex AI variant - model_vertex = "vertex_ai/gemini-3-pro-preview" - assert get_max_tokens(model_vertex) == expected Refactor the test_gemini_3_pro_preview test by using pytest.mark.parametrize to test both model variants, which reduces code duplication. tests/unittest/test_get_max_tokens.py [69-87] -def test_gemini_3_pro_preview(self, monkeypatch): - ... - # Test Google AI Studio variant - model_gemini = "gemini/gemini-3-pro-preview" - expected = 1048576 - assert get_max_tokens(model_gemini) == expected +@pytest.mark.parametrize("model", [ + "gemini/gemini-3-pro-preview", + "vertex_ai/gemini-3-pro-preview", +]) +def test_gemini_3_pro_preview(self, monkeypatch, model): + fake_settings = type("", (), { + "config": type("", (), { + "custom_model_max_tokens": 0, + "max_model_tokens": 0, + })() + })() + monkeypatch.setattr(utils, "get_settings", lambda: fake_settings) + assert get_max_tokens(model) == 1048576 - # Test Vertex AI variant - model_vertex = "vertex_ai/gemini-3-pro-preview" - assert get_max_tokens(model_vertex) == expected - [To ensure code accuracy, apply this suggestion manually] Suggestion importance[1-10]: 5 __ Why: The suggestion correctly identifies duplicated test logic and proposes a valid refactoring using pytest.mark.parametrize, which improves code maintainability and readability.	Low
Update

• Author self-review: I have reviewed the PR code suggestions, and addressed the relevant ones.

[ifox777]

/ask о чем код?

[ifox777]

/ask о чем код?

[qodo-free-for-open-source-projects]

Эта строка — не “код” приложения, а пример настройки в документации (формат TOML).

model="gemini/gemini-3-pro-preview" показывает, как в конфиге Qodo Merge зафиксировать используемую модель на gemini-3-pro-preview у провайдера gemini (в отличие от варианта выше в доке с vertex_ai/gemini-3-pro-preview).

То есть смысл: при запуске/работе Qodo Merge будет выбирать именно эту модель, а не какую-то другую по умолчанию.