fix(gemini): use thinking level for Gemini 3.5+

pyproject.toml

@@ -39,12 +39,12 @@ mistral = [
 anthropic = []
 
 gemini = [
-  "google-genai",
+  "google-genai>=1.51.0",
   "google-cloud-storage",
 ]
 
 vertexai = [
-  "google-genai",
+  "google-genai>=1.51.0",
   "google-cloud-storage",
 ]
 

src/any_llm/providers/gemini/base.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import json
+import re
 from typing import TYPE_CHECKING, Any, ClassVar, cast
 
 from typing_extensions import override
@@ -68,7 +69,26 @@
     "xhigh": 32768,
     "max": 32768,
 }
+REASONING_EFFORT_TO_THINKING_LEVELS = {
+    "minimal": types.ThinkingLevel.MINIMAL,
+    "low": types.ThinkingLevel.LOW,
+    "medium": types.ThinkingLevel.MEDIUM,
+    "high": types.ThinkingLevel.HIGH,
+    "xhigh": types.ThinkingLevel.HIGH,
+    "max": types.ThinkingLevel.HIGH,
+}
 _SUPPORTED_BATCH_ENDPOINTS = frozenset({"/v1/chat/completions"})
+_THINKING_LEVEL_MIN_GEMINI_VERSION = (3, 5)
+_GEMINI_VERSION_PATTERN = re.compile(r"(?:^|/)gemini-(\d+)(?:\.(\d+))?")
+
+
+def _uses_thinking_level(model_id: str) -> bool:
+    """Gemini 3.5 and newer reject `thinking_budget` and expect `thinking_level` instead."""
+    match = _GEMINI_VERSION_PATTERN.search(model_id.lower())
+    if match is None:
+        return False
+    major, minor = int(match.group(1)), int(match.group(2) or 0)
+    return (major, minor) >= _THINKING_LEVEL_MIN_GEMINI_VERSION
 
 
 class GoogleProvider(AnyLLM):
@@ -139,6 +159,10 @@ def _convert_completion_params(params: CompletionParams, **kwargs: Any) -> dict[
         if params.reasoning_effort != "auto":
             if params.reasoning_effort is None or params.reasoning_effort == "none":
                 kwargs["thinking_config"] = types.ThinkingConfig(include_thoughts=False)
+            elif _uses_thinking_level(params.model_id):
+                kwargs["thinking_config"] = types.ThinkingConfig(
+                    include_thoughts=True, thinking_level=REASONING_EFFORT_TO_THINKING_LEVELS[params.reasoning_effort]
+                )
             else:
                 kwargs["thinking_config"] = types.ThinkingConfig(
                     include_thoughts=True, thinking_budget=REASONING_EFFORT_TO_THINKING_BUDGETS[params.reasoning_effort]

tests/unit/providers/test_gemini_provider.py

@@ -598,6 +598,52 @@ async def test_completion_with_custom_reasoning_effort(reasoning_effort: Reasoni
             )
 
 
+@pytest.mark.parametrize(
+    ("model_id", "reasoning_effort", "expected_level"),
+    [
+        ("gemini-3.5-flash", "xhigh", types.ThinkingLevel.HIGH),
+        ("gemini-3.5-flash", "max", types.ThinkingLevel.HIGH),
+        ("gemini-3.5-pro", "low", types.ThinkingLevel.LOW),
+        ("models/gemini-3.5-flash", "medium", types.ThinkingLevel.MEDIUM),
+        ("gemini-3.10-flash", "minimal", types.ThinkingLevel.MINIMAL),
+        ("gemini-4-pro", "high", types.ThinkingLevel.HIGH),
+    ],
+)
+def test_new_gemini_models_use_thinking_level(
+    model_id: str, reasoning_effort: ReasoningEffort, expected_level: types.ThinkingLevel
+) -> None:
+    result = GoogleProvider._convert_completion_params(
+        CompletionParams(
+            model_id=model_id, messages=[{"role": "user", "content": "Hello"}], reasoning_effort=reasoning_effort
+        ),
+        provider_name="gemini",
+    )
+
+    assert result["config"].thinking_config == types.ThinkingConfig(
+        include_thoughts=True, thinking_level=expected_level
+    )
+
+
+@pytest.mark.parametrize(
+    "model_id",
+    [
+        "gemini-3.0-flash",
+        "gemini-3.4-flash",
+        "gemini-3-pro-preview",
+        "gemini-2.5-flash",
+        "gemini-pro",
+        "projects/p/locations/l/publishers/google/models/gemini-3-pro",
+    ],
+)
+def test_older_gemini_models_keep_thinking_budget(model_id: str) -> None:
+    result = GoogleProvider._convert_completion_params(
+        CompletionParams(model_id=model_id, messages=[{"role": "user", "content": "Hello"}], reasoning_effort="high"),
+        provider_name="gemini",
+    )
+
+    assert result["config"].thinking_config == types.ThinkingConfig(include_thoughts=True, thinking_budget=24576)
+
+
 @pytest.mark.asyncio
 async def test_completion_with_max_tokens_conversion() -> None:
     """Test that max_tokens parameter gets converted to max_output_tokens."""