diff --git a/litellm/llms/brave/search/transformation.py b/litellm/llms/brave/search/transformation.py
index 1f78f8d154c..0b8ebe9792b 100644
--- a/litellm/llms/brave/search/transformation.py
+++ b/litellm/llms/brave/search/transformation.py
@@ -175,6 +175,8 @@ class BraveSearchConfig(BaseSearchConfig):
         - max_results → count
         - search_domain_filter → q (append domain filters)
         - country → country
+        - start_date + end_date → freshness (combined as "YYYY-MM-DDtoYYYY-MM-DD";
+        only sent when both are provided — Brave's custom range requires both)
         - max_tokens_per_page → (not applicable, ignored)
 
         All other Brave Search API-specific parameters are passed through as-is.
@@ -190,36 +192,48 @@ class BraveSearchConfig(BaseSearchConfig):
             # Brave Search API only supports single string queries
             query = " ".join(query)
 
+        remaining = dict(optional_params)
+
         request_data: Final[BraveSearchRequest] = {
             "q": query,
         }
 
         # Only include "include_fetch_metadata" if it is not explicitly set to False
         # This parameter results (more often than not) in a timestamp which we can use for last_updated
-        if "include_fetch_metadata" in optional_params and optional_params["include_fetch_metadata"] is False:
+        if remaining.pop("include_fetch_metadata", None) is False:
             request_data["include_fetch_metadata"] = False
         else:
             request_data["include_fetch_metadata"] = True
 
         # Transform unified spec parameters to Brave Search API format
-        if "max_results" in optional_params:
+        if "max_results" in remaining:
             # Brave Search API supports 1-20 results per /web/search request
-            num_results: Final = min(optional_params["max_results"], 20)
+            num_results: Final = min(remaining.pop("max_results"), 20)
             request_data["count"] = num_results
 
         if "search_domain_filter" in optional_params:
             # Convert to multiple "site:domain" clauses, joined by OR
-            domains: Final = optional_params["search_domain_filter"]
+            domains: Final = remaining.pop("search_domain_filter")
             if isinstance(domains, list) and len(domains) > 0:
                 request_data["q"] = self._append_domain_filters(request_data["q"], domains)
 
+        start_date = remaining.pop("start_date", None)
+        end_date = remaining.pop("end_date", None)
+        if start_date and end_date:
+            request_data["freshness"] = f"{start_date}to{end_date}"
+
+        if "country" in remaining:
+            request_data["country"] = remaining.pop("country")
+
+        # Not applicable therefore popped and ignored
+        if "max_tokens_per_page" in remaining:
+            remaining.pop("max_tokens_per_page")
+
         # Convert to dict before dynamic key assignments
         result_data: Final = dict(request_data)
 
-        # Pass through all other parameters as-is
-        for param, value in optional_params.items():
-            if param not in self.get_supported_perplexity_optional_params() and param not in result_data:
-                result_data[param] = value
+        # Pass through anything the function did not explicitly consume
+        result_data.update(remaining)
 
         # Store params in special key for URL building (Brave Search API uses GET not POST)
         # Return a wrapper dict that stores params for get_complete_url to use
diff --git a/litellm/llms/linkup/search/transformation.py b/litellm/llms/linkup/search/transformation.py
index 8ecd5a24c08..40cb7c56a7e 100644
--- a/litellm/llms/linkup/search/transformation.py
+++ b/litellm/llms/linkup/search/transformation.py
@@ -104,6 +104,8 @@ class LinkupSearchConfig(BaseSearchConfig):
         - max_results -> maxResults
         - search_domain_filter -> includeDomains
         - country -> (not directly supported)
+        - start_date -> fromDate
+        - end_date -> toDate
         - max_tokens_per_page -> (not applicable)
 
         All other Linkup-specific parameters are passed through as-is.
@@ -119,26 +121,36 @@ class LinkupSearchConfig(BaseSearchConfig):
             # Linkup only supports single string queries, join with spaces
             query = " ".join(query)
 
+        # Copy for passthrough data (Done this way to avoid having to change / add to Perplexity unified spec parameters)
+        remaining = dict(optional_params)
+
         request_data: Final[LinkupSearchRequest] = {
             "q": query,
-            "depth": optional_params.get("depth", "standard"),
-            "outputType": optional_params.get("outputType", "searchResults"),
+            "depth": remaining.pop("depth", "standard"),
+            "outputType": remaining.pop("outputType", "searchResults"),
         }
 
         # Transform Perplexity unified spec parameters to Linkup format
-        if "max_results" in optional_params:
-            request_data["maxResults"] = optional_params["max_results"]
+        if "max_results" in remaining:
+            request_data["maxResults"] = remaining.pop("max_results")
 
-        if "search_domain_filter" in optional_params:
-            request_data["includeDomains"] = optional_params["search_domain_filter"]
+        if "search_domain_filter" in remaining:
+            request_data["includeDomains"] = remaining.pop("search_domain_filter")
+
+        if "start_date" in remaining:
+            request_data["fromDate"] = remaining.pop("start_date")
+
+        if "end_date" in remaining:
+            request_data["toDate"] = remaining.pop("end_date")
+
+        if "max_tokens_per_page" in remaining:
+            remaining.pop("max_tokens_per_page")
 
         # Convert to dict before dynamic key assignments
         result_data: Final = dict(request_data)
 
-        # pass through all other parameters as-is
-        for param, value in optional_params.items():
-            if param not in self.get_supported_perplexity_optional_params() and param not in result_data:
-                result_data[param] = value
+        # Pass through any unhandled data
+        result_data.update(remaining)
 
         return result_data
 
@@ -189,7 +201,7 @@ class LinkupSearchConfig(BaseSearchConfig):
             elif result_type == "image":
                 # For image results, use the URL as both title and snippet if name not provided
                 search_result = SearchResult(
-                    title=result.get("name", result.get("url", "")),
+                    title=result.get("name") or result.get("url", ""),
                     url=result.get("url", ""),
                     snippet=result.get("content", ""),
                     date=None,
diff --git a/litellm/search/main.py b/litellm/search/main.py
index b2dd51799a1..20fa6046553 100644
--- a/litellm/search/main.py
+++ b/litellm/search/main.py
@@ -29,6 +29,8 @@ def _build_search_optional_params(
     search_domain_filter: list[str] | None = None,
     max_tokens_per_page: int | None = None,
     country: str | None = None,
+    start_date: str | None = None,
+    end_date: str | None = None,
 ) -> dict[str, Any]:
     """
     Helper function to build optional_params dict from Perplexity Search API parameters.
@@ -38,6 +40,8 @@ def _build_search_optional_params(
         search_domain_filter: List of domains to filter (max 20)
         max_tokens_per_page: Max tokens per page
         country: Country code filter
+        start_date: Start date for results (YYYY-MM-DD)
+        end_date: End date for results (YYYY-MM-DD)
 
     Returns:
         Dict with non-None optional parameters
@@ -52,6 +56,10 @@ def _build_search_optional_params(
         optional_params["max_tokens_per_page"] = max_tokens_per_page
     if country is not None:
         optional_params["country"] = country
+    if start_date is not None:
+        optional_params["start_date"] = start_date
+    if end_date is not None:
+        optional_params["end_date"] = end_date
 
     return optional_params
 
@@ -64,6 +72,8 @@ async def asearch(
     search_domain_filter: list[str] | None = None,
     max_tokens_per_page: int | None = None,
     country: str | None = None,
+    start_date: str | None = None,
+    end_date: str | None = None,
     api_key: str | None = None,
     api_base: str | None = None,
     timeout: float | httpx.Timeout | None = None,
@@ -80,6 +90,8 @@ async def asearch(
         search_domain_filter: Optional list of domains to filter (max 20)
         max_tokens_per_page: Optional max tokens per page, default 1024
         country: Optional country code filter (e.g., 'US', 'GB', 'DE')
+        start_date: Optional start date for results (YYYY-MM-DD)
+        end_date: Optional end date for results (YYYY-MM-DD)
         api_key: Optional API key
         api_base: Optional API base URL
         timeout: Optional timeout
@@ -128,6 +140,8 @@ async def asearch(
             search_domain_filter=search_domain_filter,
             max_tokens_per_page=max_tokens_per_page,
             country=country,
+            start_date=start_date,
+            end_date=end_date,
             api_key=api_key,
             api_base=api_base,
             timeout=timeout,
@@ -167,6 +181,8 @@ def search(
     search_domain_filter: list[str] | None = None,
     max_tokens_per_page: int | None = None,
     country: str | None = None,
+    start_date: str | None = None,
+    end_date: str | None = None,
     api_key: str | None = None,
     api_base: str | None = None,
     timeout: float | httpx.Timeout | None = None,
@@ -183,6 +199,8 @@ def search(
         search_domain_filter: Optional list of domains to filter (max 20)
         max_tokens_per_page: Optional max tokens per page, default 1024
         country: Optional country code filter (e.g., 'US', 'GB', 'DE')
+        start_date: Optional start date for results (YYYY-MM-DD)
+        end_date: Optional end date for results (YYYY-MM-DD)
         api_key: Optional API key
         api_base: Optional API base URL
         timeout: Optional timeout
@@ -255,6 +273,8 @@ def search(
             search_domain_filter=search_domain_filter,
             max_tokens_per_page=max_tokens_per_page,
             country=country,
+            start_date=start_date,
+            end_date=end_date,
         )
 
         # Filter out internal LiteLLM parameters from kwargs
diff --git a/tests/test_litellm/llms/brave/search/test_brave_search_transformation.py b/tests/test_litellm/llms/brave/search/test_brave_search_transformation.py
new file mode 100644
index 00000000000..d21b297eb61
--- /dev/null
+++ b/tests/test_litellm/llms/brave/search/test_brave_search_transformation.py
@@ -0,0 +1,279 @@
+from unittest.mock import Mock
+
+import pytest
+
+from litellm.llms.brave.search.transformation import BraveSearchConfig, to_yyyy_mm_dd
+
+
+def _config() -> BraveSearchConfig:
+    return BraveSearchConfig()
+
+
+def _resp(payload: dict, params: dict | None = None) -> Mock:
+    """Mock httpx.Response with the .json() and .request.url.params shape
+    Brave's transform_search_response actually reads."""
+    r = Mock()
+    r.json.return_value = payload
+    req = Mock()
+    req.url = Mock()
+    req.url.params = params or {}
+    r.request = req
+    return r
+
+
+def _result(**overrides):
+    base = {
+        "title": "Test Title",
+        "url": "https://example.com",
+        "description": "Test description",
+        "page_age": None,
+        "age": None,
+        "fetched_content_timestamp": None,
+    }
+    return {**base, **overrides}
+
+
+# --- ui_friendly_name / get_http_method ---
+
+
+def test_ui_friendly_name():
+    assert _config().ui_friendly_name() == "Brave Search"
+
+
+def test_get_http_method_is_get():
+    """Brave's /web/search endpoint takes query params, not a JSON body."""
+    assert _config().get_http_method() == "GET"
+
+
+# --- validate_environment ---
+
+
+def test_validate_environment_with_explicit_key():
+    headers = _config().validate_environment({}, api_key="explicit-key")
+    assert headers["X-Subscription-Token"] == "explicit-key"
+    assert headers["Accept"] == "application/json"
+    assert headers["Accept-Encoding"] == "gzip"
+    assert headers["Content-Type"] == "application/json"
+
+
+def test_validate_environment_reads_env_key(monkeypatch: pytest.MonkeyPatch):
+    monkeypatch.setenv("BRAVE_API_KEY", "env-key")
+    assert _config().validate_environment({})["X-Subscription-Token"] == "env-key"
+
+
+def test_validate_environment_missing_key_raises(monkeypatch: pytest.MonkeyPatch):
+    monkeypatch.delenv("BRAVE_API_KEY", raising=False)
+    with pytest.raises(ValueError, match="BRAVE_API_KEY"):
+        _config().validate_environment({})
+
+
+# --- get_complete_url ---
+
+
+def test_get_complete_url_default_base(monkeypatch: pytest.MonkeyPatch):
+    monkeypatch.delenv("BRAVE_API_BASE", raising=False)
+    url = _config().get_complete_url(None, {}, data={"_brave_params": {"q": "test"}})
+    assert url.startswith("https://api.search.brave.com/res/v1/web/search?")
+    assert "q=test" in url
+
+
+def test_get_complete_url_reads_env_base(monkeypatch: pytest.MonkeyPatch):
+    monkeypatch.setenv("BRAVE_API_BASE", "https://env-base.local/search")
+    url = _config().get_complete_url(None, {}, data={"_brave_params": {"q": "test"}})
+    assert url.startswith("https://env-base.local/search?")
+
+
+def test_get_complete_url_without_brave_params_returns_bare_base(monkeypatch: pytest.MonkeyPatch):
+    monkeypatch.delenv("BRAVE_API_BASE", raising=False)
+    assert _config().get_complete_url(None, {}, data=None) == "https://api.search.brave.com/res/v1/web/search"
+
+
+# --- transform_search_request ---
+
+
+def test_transform_search_request_joins_list_query():
+    data = _config().transform_search_request(["foo", "bar"], {})
+    assert data["_brave_params"]["q"] == "foo bar"
+
+
+def test_transform_search_request_max_results_clamped_to_20():
+    """Unlike Nimble, Brave's /web/search hard-caps at 20 results per request."""
+    data = _config().transform_search_request("q", {"max_results": 500})
+    assert data["_brave_params"]["count"] == 20
+
+
+def test_transform_search_request_max_results_under_cap_passes_through():
+    data = _config().transform_search_request("q", {"max_results": 5})
+    assert data["_brave_params"]["count"] == 5
+
+
+def test_transform_search_request_appends_domain_filters_as_site_clauses():
+    data = _config().transform_search_request("q", {"search_domain_filter": ["arxiv.org", "nature.com"]})
+    assert data["_brave_params"]["q"] == "(q) AND (site:arxiv.org OR site:nature.com)"
+
+
+def test_transform_search_request_empty_domain_filter_leaves_query_unchanged():
+    data = _config().transform_search_request("q", {"search_domain_filter": []})
+    assert data["_brave_params"]["q"] == "q"
+
+
+def test_transform_search_request_drops_max_tokens_per_page():
+    """No Brave equivalent — must not leak through as an unrecognized param."""
+    assert "max_tokens_per_page" not in _config().transform_search_request("q", {"max_tokens_per_page": 1024})[
+        "_brave_params"
+    ]
+
+
+def test_transform_search_request_country_is_forwarded():
+    """Regression test: the docstring always claimed country -> country, but
+    there was previously no code path that actually implemented it — it
+    relied on shared-list passthrough, which excluded it."""
+    data = _config().transform_search_request("q", {"country": "US"})
+    assert data["_brave_params"]["country"] == "US"
+
+
+def test_transform_search_request_passes_through_unhandled_kwargs():
+    data = _config().transform_search_request("q", {"safesearch": "strict"})
+    assert data["_brave_params"]["safesearch"] == "strict"
+
+
+@pytest.mark.parametrize(
+    "optional_params",
+    [
+        {},  # absent entirely
+        {"include_fetch_metadata": True},  # explicitly True
+    ],
+)
+def test_transform_search_request_include_fetch_metadata_defaults_true(optional_params):
+    data = _config().transform_search_request("q", optional_params)
+    assert data["_brave_params"]["include_fetch_metadata"] is True
+
+
+def test_transform_search_request_include_fetch_metadata_explicit_false_is_respected():
+    data = _config().transform_search_request("q", {"include_fetch_metadata": False})
+    assert data["_brave_params"]["include_fetch_metadata"] is False
+
+
+# --- transform_search_request: date filtering ---
+
+
+def test_transform_search_request_date_range_maps_to_freshness():
+    data = _config().transform_search_request("q", {"start_date": "2022-04-01", "end_date": "2022-07-30"})
+    assert data["_brave_params"]["freshness"] == "2022-04-01to2022-07-30"
+
+
+def test_transform_search_request_only_start_date_is_dropped():
+    """Brave's custom range needs both bounds; there's no correct open-ended equivalent."""
+    data = _config().transform_search_request("q", {"start_date": "2022-04-01"})
+    params = data["_brave_params"]
+    assert "freshness" not in params
+    assert "start_date" not in params  # must not leak through raw either
+
+
+def test_transform_search_request_only_end_date_is_dropped():
+    data = _config().transform_search_request("q", {"end_date": "2022-07-30"})
+    params = data["_brave_params"]
+    assert "freshness" not in params
+    assert "end_date" not in params
+
+
+def test_transform_search_request_no_dates_omits_freshness():
+    data = _config().transform_search_request("q", {"max_results": 5})
+    assert "freshness" not in data["_brave_params"]
+
+
+def test_transform_search_request_explicit_freshness_not_clobbered_by_absent_dates():
+    """A caller-supplied native `freshness` preset (e.g. 'pw') should survive
+    when start_date/end_date aren't given at all."""
+    data = _config().transform_search_request("q", {"freshness": "pw"})
+    assert data["_brave_params"]["freshness"] == "pw"
+
+
+# --- transform_search_response ---
+
+
+def test_transform_search_response_extracts_basic_fields():
+    resp = _resp({"web": {"results": [_result()]}})
+    result = _config().transform_search_response(resp, logging_obj=Mock()).results[0]
+    assert result.title == "Test Title"
+    assert result.url == "https://example.com"
+    assert result.snippet == "Test description"
+
+
+def test_transform_search_response_date_from_page_age():
+    resp = _resp({"web": {"results": [_result(page_age="2024-01-15")]}})
+    result = _config().transform_search_response(resp, logging_obj=Mock()).results[0]
+    assert result.date == "2024-01-15"
+
+
+def test_transform_search_response_date_falls_back_to_age_when_page_age_absent():
+    resp = _resp({"web": {"results": [_result(page_age=None, age="2024-01-15")]}})
+    result = _config().transform_search_response(resp, logging_obj=Mock()).results[0]
+    assert result.date == "2024-01-15"
+
+
+def test_transform_search_response_last_updated_from_fetched_content_timestamp():
+    resp = _resp({"web": {"results": [_result(fetched_content_timestamp="1705334400")]}})
+    result = _config().transform_search_response(resp, logging_obj=Mock()).results[0]
+    assert result.last_updated == "2024-01-15"
+
+
+def test_transform_search_response_zero_hits():
+    resp = _resp({"web": {"results": []}})
+    assert _config().transform_search_response(resp, logging_obj=Mock()).results == []
+
+
+def test_transform_search_response_result_filter_limits_sections():
+    resp = _resp(
+        {
+            "web": {"results": [_result(title="web result")]},
+            "news": {"results": [_result(title="news result")]},
+        },
+        params={"result_filter": "news"},
+    )
+    titles = [r.title for r in _config().transform_search_response(resp, logging_obj=Mock()).results]
+    assert titles == ["news result"]
+
+
+def test_transform_search_response_no_result_filter_includes_all_sections():
+    resp = _resp(
+        {
+            "web": {"results": [_result(title="web result")]},
+            "news": {"results": [_result(title="news result")]},
+        },
+    )
+    titles = {r.title for r in _config().transform_search_response(resp, logging_obj=Mock()).results}
+    assert titles == {"web result", "news result"}
+
+
+def test_transform_search_response_max_results_limits_across_combined_sections():
+    """count doesn't limit Brave's own per-section results server-side, so
+    LiteLLM has to truncate across sections manually."""
+    resp = _resp(
+        {
+            "web": {"results": [_result(title=f"web-{i}") for i in range(5)]},
+            "news": {"results": [_result(title=f"news-{i}") for i in range(5)]},
+        },
+        params={"count": "3"},
+    )
+    results = _config().transform_search_response(resp, logging_obj=Mock()).results
+    assert len(results) == 3
+
+
+# --- to_yyyy_mm_dd helper ---
+
+
+@pytest.mark.parametrize(
+    "value,expected",
+    [
+        (None, None),
+        ("", None),
+        ("not-a-date", None),
+        ("2024-01-15", "2024-01-15"),
+        ("2024/01/15", "2024-01-15"),
+        (1705334400, "2024-01-15"),  # unix seconds
+        (1705334400000, "2024-01-15"),  # unix milliseconds
+    ],
+)
+def test_to_yyyy_mm_dd(value, expected):
+    assert to_yyyy_mm_dd(value) == expected
\ No newline at end of file
diff --git a/tests/test_litellm/llms/linkup/search/test_linkup_search_transformation.py b/tests/test_litellm/llms/linkup/search/test_linkup_search_transformation.py
new file mode 100644
index 00000000000..b172c5f3deb
--- /dev/null
+++ b/tests/test_litellm/llms/linkup/search/test_linkup_search_transformation.py
@@ -0,0 +1,220 @@
+from unittest.mock import Mock
+
+import pytest
+
+from litellm.llms.linkup.search.transformation import LinkupSearchConfig
+
+
+def _config() -> LinkupSearchConfig:
+    return LinkupSearchConfig()
+
+
+def _resp(payload: dict) -> Mock:
+    r = Mock()
+    r.json.return_value = payload
+    return r
+
+
+def _result(**overrides):
+    base = {
+        "type": "text",
+        "name": "Test Title",
+        "url": "https://example.com",
+        "content": "Test content",
+    }
+    return {**base, **overrides}
+
+
+# --- ui_friendly_name ---
+
+
+def test_ui_friendly_name():
+    assert _config().ui_friendly_name() == "Linkup"
+
+
+# --- validate_environment ---
+
+
+def test_validate_environment_with_explicit_key():
+    headers = _config().validate_environment({}, api_key="explicit-key")
+    assert headers["Authorization"] == "Bearer explicit-key"
+    assert headers["Content-Type"] == "application/json"
+
+
+def test_validate_environment_reads_env_key(monkeypatch: pytest.MonkeyPatch):
+    monkeypatch.setenv("LINKUP_API_KEY", "env-key")
+    assert _config().validate_environment({})["Authorization"] == "Bearer env-key"
+
+
+def test_validate_environment_missing_key_raises(monkeypatch: pytest.MonkeyPatch):
+    monkeypatch.delenv("LINKUP_API_KEY", raising=False)
+    with pytest.raises(ValueError, match="LINKUP_API_KEY"):
+        _config().validate_environment({})
+
+
+# --- get_complete_url ---
+
+
+def test_get_complete_url_default_base(monkeypatch: pytest.MonkeyPatch):
+    monkeypatch.delenv("LINKUP_API_BASE", raising=False)
+    assert _config().get_complete_url(None, {}) == "https://api.linkup.so/v1/search"
+
+
+def test_get_complete_url_reads_env_base(monkeypatch: pytest.MonkeyPatch):
+    monkeypatch.setenv("LINKUP_API_BASE", "https://env-base.local/v1")
+    assert _config().get_complete_url(None, {}) == "https://env-base.local/v1/search"
+
+
+def test_get_complete_url_does_not_duplicate_search_suffix():
+    assert _config().get_complete_url("https://self-hosted.local/v1/search", {}) == "https://self-hosted.local/v1/search"
+
+
+def test_get_complete_url_appends_search_when_missing():
+    assert _config().get_complete_url("https://self-hosted.local/v1", {}) == "https://self-hosted.local/v1/search"
+
+
+# --- transform_search_request: query / defaults ---
+
+
+def test_transform_search_request_joins_list_query():
+    assert _config().transform_search_request(["foo", "bar"], {})["q"] == "foo bar"
+
+
+def test_transform_search_request_string_query_unchanged():
+    assert _config().transform_search_request("foo bar", {})["q"] == "foo bar"
+
+
+def test_transform_search_request_defaults_depth_and_output_type():
+    data = _config().transform_search_request("q", {})
+    assert data["depth"] == "standard"
+    assert data["outputType"] == "searchResults"
+
+
+def test_transform_search_request_respects_explicit_depth_and_output_type():
+    data = _config().transform_search_request("q", {"depth": "deep", "outputType": "sourcedAnswer"})
+    assert data["depth"] == "deep"
+    assert data["outputType"] == "sourcedAnswer"
+
+
+# --- transform_search_request: unified param mapping ---
+
+
+def test_transform_search_request_max_results_maps_to_maxResults():
+    data = _config().transform_search_request("q", {"max_results": 10})
+    assert data["maxResults"] == 10
+    assert "max_results" not in data
+
+
+def test_transform_search_request_search_domain_filter_maps_to_includeDomains():
+    data = _config().transform_search_request("q", {"search_domain_filter": ["arxiv.org", "nature.com"]})
+    assert data["includeDomains"] == ["arxiv.org", "nature.com"]
+    assert "search_domain_filter" not in data
+
+
+def test_transform_search_request_drops_max_tokens_per_page():
+    """No Linkup equivalent — must not leak through as an unrecognized param."""
+    assert "max_tokens_per_page" not in _config().transform_search_request("q", {"max_tokens_per_page": 1024})
+
+
+def test_transform_search_request_country_has_no_native_equivalent():
+    """Linkup has no native country filter; per the docstring this is intentionally unmapped —
+    it should pass through raw rather than silently vanish, since Linkup's real API will just
+    ignore an unrecognized field."""
+    data = _config().transform_search_request("q", {"country": "US"})
+    assert data.get("country") == "US"
+
+
+# --- transform_search_request: date filtering ---
+
+
+def test_transform_search_request_maps_start_date_to_from_date():
+    data = _config().transform_search_request("q", {"start_date": "1999-03-20"})
+    assert data["fromDate"] == "1999-03-20"
+    assert "start_date" not in data
+
+
+def test_transform_search_request_maps_end_date_to_to_date():
+    data = _config().transform_search_request("q", {"end_date": "1999-04-20"})
+    assert data["toDate"] == "1999-04-20"
+    assert "end_date" not in data
+
+
+def test_transform_search_request_date_range_together():
+    data = _config().transform_search_request(
+        "q", {"start_date": "1999-03-20", "end_date": "1999-04-20"}
+    )
+    assert data["fromDate"] == "1999-03-20"
+    assert data["toDate"] == "1999-04-20"
+
+
+def test_transform_search_request_without_dates_omits_both():
+    data = _config().transform_search_request("q", {"max_results": 5})
+    assert "fromDate" not in data
+    assert "toDate" not in data
+
+
+# --- transform_search_request: passthrough ---
+
+
+def test_transform_search_request_passes_through_unhandled_kwargs():
+    """A param this function doesn't explicitly handle should still reach Linkup unchanged."""
+    data = _config().transform_search_request("q", {"includeImages": True})
+    assert data["includeImages"] is True
+
+
+def test_transform_search_request_does_not_duplicate_consumed_params():
+    """Regression test: a translated param (start_date -> fromDate) must not
+    also leak through raw via the passthrough path."""
+    data = _config().transform_search_request(
+        "q", {"start_date": "1999-03-20", "max_results": 5, "includeImages": True}
+    )
+    assert "start_date" not in data
+    assert "max_results" not in data
+    assert data["fromDate"] == "1999-03-20"
+    assert data["maxResults"] == 5
+    assert data["includeImages"] is True
+
+
+# --- transform_search_response ---
+
+
+def test_transform_search_response_text_result_fields():
+    resp = _resp({"results": [_result()]})
+    result = _config().transform_search_response(resp, logging_obj=Mock()).results[0]
+    assert result.title == "Test Title"
+    assert result.url == "https://example.com"
+    assert result.snippet == "Test content"
+    assert result.date is None
+    assert result.last_updated is None
+
+
+def test_transform_search_response_image_result_falls_back_to_url_for_title():
+    resp = _resp({"results": [_result(type="image", name="", url="https://example.com/img.png")]})
+    result = _config().transform_search_response(resp, logging_obj=Mock()).results[0]
+    assert result.title == "https://example.com/img.png"
+
+
+def test_transform_search_response_image_result_uses_name_when_present():
+    resp = _resp({"results": [_result(type="image", name="Alt text")]})
+    result = _config().transform_search_response(resp, logging_obj=Mock()).results[0]
+    assert result.title == "Alt text"
+
+
+def test_transform_search_response_preserves_order():
+    resp = _resp({"results": [_result(name=n) for n in ("first", "second", "third")]})
+    titles = [r.title for r in _config().transform_search_response(resp, logging_obj=Mock()).results]
+    assert titles == ["first", "second", "third"]
+
+
+def test_transform_search_response_zero_hits():
+    resp = _resp({"results": []})
+    assert _config().transform_search_response(resp, logging_obj=Mock()).results == []
+
+
+def test_transform_search_response_unknown_result_type_is_skipped():
+    """A result type that's neither "text" nor "image" (e.g. a future Linkup
+    addition) shouldn't crash the call; it's just silently omitted."""
+    resp = _resp({"results": [_result(type="video"), _result()]})
+    results = _config().transform_search_response(resp, logging_obj=Mock()).results
+    assert len(results) == 1
+    assert results[0].title == "Test Title"
\ No newline at end of file
diff --git a/tests/test_litellm/llms/nimble/search/test_nimble_search_transformation.py b/tests/test_litellm/llms/nimble/search/test_nimble_search_transformation.py
index d6292c9cf3e..dc1e8ab5625 100644
--- a/tests/test_litellm/llms/nimble/search/test_nimble_search_transformation.py
+++ b/tests/test_litellm/llms/nimble/search/test_nimble_search_transformation.py
@@ -100,6 +100,14 @@ def test_transform_search_request_max_results_is_not_clamped():
 def test_transform_search_request_uppercases_country():
     assert _config().transform_search_request("q", {"country": "us"})["country"] == "US"
 
+def test_transform_search_request_date_range_passes_through_native_names():
+    """Nimble already uses the unified spec's own field names for date filtering."""
+    data = _config().transform_search_request(
+        "q", {"start_date": "1999-03-20", "end_date": "1999-04-20"}
+    )
+    assert data["start_date"] == "1999-03-20"
+    assert data["end_date"] == "1999-04-20"
+
 
 def test_transform_search_request_drops_max_tokens_per_page():
     assert "max_tokens_per_page" not in _config().transform_search_request("q", {"max_tokens_per_page": 1024})
diff --git a/tests/test_litellm/llms/tavily/search/test_tavily_search_transformation.py b/tests/test_litellm/llms/tavily/search/test_tavily_search_transformation.py
new file mode 100644
index 00000000000..80ada94d22c
--- /dev/null
+++ b/tests/test_litellm/llms/tavily/search/test_tavily_search_transformation.py
@@ -0,0 +1,180 @@
+from unittest.mock import Mock
+
+import pytest
+
+from litellm.llms.tavily.search.transformation import TavilySearchConfig
+
+
+def _config() -> TavilySearchConfig:
+    return TavilySearchConfig()
+
+
+def _resp(payload: dict) -> Mock:
+    r = Mock()
+    r.json.return_value = payload
+    return r
+
+
+def _result(**overrides):
+    base = {
+        "title": "Test Title",
+        "url": "https://example.com",
+        "content": "Test content",
+    }
+    return {**base, **overrides}
+
+
+# --- ui_friendly_name ---
+
+
+def test_ui_friendly_name():
+    assert _config().ui_friendly_name() == "Tavily"
+
+
+# --- validate_environment ---
+
+
+def test_validate_environment_with_explicit_key():
+    headers = _config().validate_environment({}, api_key="explicit-key")
+    assert headers["Authorization"] == "Bearer explicit-key"
+    assert headers["Content-Type"] == "application/json"
+
+
+def test_validate_environment_reads_env_key(monkeypatch: pytest.MonkeyPatch):
+    monkeypatch.setenv("TAVILY_API_KEY", "env-key")
+    assert _config().validate_environment({})["Authorization"] == "Bearer env-key"
+
+
+def test_validate_environment_missing_key_raises(monkeypatch: pytest.MonkeyPatch):
+    monkeypatch.delenv("TAVILY_API_KEY", raising=False)
+    with pytest.raises(ValueError, match="TAVILY_API_KEY"):
+        _config().validate_environment({})
+
+
+# --- get_complete_url ---
+
+
+def test_get_complete_url_default_base(monkeypatch: pytest.MonkeyPatch):
+    monkeypatch.delenv("TAVILY_API_BASE", raising=False)
+    assert _config().get_complete_url(None, {}) == "https://api.tavily.com/search"
+
+
+def test_get_complete_url_reads_env_base(monkeypatch: pytest.MonkeyPatch):
+    monkeypatch.setenv("TAVILY_API_BASE", "https://env-base.local")
+    assert _config().get_complete_url(None, {}) == "https://env-base.local/search"
+
+
+def test_get_complete_url_does_not_duplicate_search_suffix():
+    assert _config().get_complete_url("https://self-hosted.local/search", {}) == "https://self-hosted.local/search"
+
+
+# --- transform_search_request: query ---
+
+
+def test_transform_search_request_joins_list_query():
+    assert _config().transform_search_request(["foo", "bar"], {})["query"] == "foo bar"
+
+
+def test_transform_search_request_string_query_unchanged():
+    assert _config().transform_search_request("foo bar", {})["query"] == "foo bar"
+
+
+# --- transform_search_request: unified param mapping ---
+
+
+def test_transform_search_request_max_results_passes_through():
+    data = _config().transform_search_request("q", {"max_results": 10})
+    assert data["max_results"] == 10
+
+
+def test_transform_search_request_search_domain_filter_maps_to_include_domains():
+    data = _config().transform_search_request("q", {"search_domain_filter": ["arxiv.org", "nature.com"]})
+    assert data["include_domains"] == ["arxiv.org", "nature.com"]
+    assert "search_domain_filter" not in data
+
+
+def test_transform_search_request_lowercases_country():
+    data = _config().transform_search_request("q", {"country": "US"})
+    assert data["country"] == "us"
+
+
+def test_transform_search_request_drops_max_tokens_per_page():
+    """No Tavily equivalent — must not leak through as an unrecognized param."""
+    assert "max_tokens_per_page" not in _config().transform_search_request("q", {"max_tokens_per_page": 1024})
+
+
+# --- transform_search_request: date filtering (native passthrough) ---
+
+
+def test_transform_search_request_date_range_passes_through_native_names():
+    """Tavily already uses the unified spec's own field names for date filtering."""
+    data = _config().transform_search_request(
+        "q", {"start_date": "1999-03-20", "end_date": "1999-04-20"}
+    )
+    assert data["start_date"] == "1999-03-20"
+    assert data["end_date"] == "1999-04-20"
+
+
+def test_transform_search_request_single_date_passes_through():
+    """Unlike Brave, Tavily has no combined-range requirement — a lone bound is valid."""
+    data = _config().transform_search_request("q", {"start_date": "1999-03-20"})
+    assert data["start_date"] == "1999-03-20"
+    assert "end_date" not in data
+
+
+def test_transform_search_request_without_dates_omits_both():
+    data = _config().transform_search_request("q", {"max_results": 5})
+    assert "start_date" not in data
+    assert "end_date" not in data
+
+
+# --- transform_search_request: passthrough ---
+
+
+def test_transform_search_request_passes_through_unhandled_kwargs():
+    """Native Tavily-specific params (topic, search_depth, time_range, etc.)
+    aren't explicitly mapped, but should still reach the request unchanged."""
+    data = _config().transform_search_request("q", {"topic": "news", "search_depth": "advanced"})
+    assert data["topic"] == "news"
+    assert data["search_depth"] == "advanced"
+
+
+def test_transform_search_request_does_not_duplicate_consumed_params():
+    """Regression test: a translated param (search_domain_filter -> include_domains)
+    must not also leak through raw via the passthrough path."""
+    data = _config().transform_search_request(
+        "q", {"search_domain_filter": ["arxiv.org"], "max_results": 5}
+    )
+    assert "search_domain_filter" not in data
+    assert data["include_domains"] == ["arxiv.org"]
+    assert data["max_results"] == 5
+
+
+# --- transform_search_response ---
+
+
+def test_transform_search_response_extracts_basic_fields():
+    resp = _resp({"results": [_result()]})
+    result = _config().transform_search_response(resp, logging_obj=Mock()).results[0]
+    assert result.title == "Test Title"
+    assert result.url == "https://example.com"
+    assert result.snippet == "Test content"
+
+
+def test_transform_search_response_date_and_last_updated_are_none():
+    """Tavily's response has no date/last_updated equivalent."""
+    resp = _resp({"results": [_result()]})
+    result = _config().transform_search_response(resp, logging_obj=Mock()).results[0]
+    assert result.date is None
+    assert result.last_updated is None
+
+
+def test_transform_search_response_zero_hits():
+    resp = _resp({"results": []})
+    assert _config().transform_search_response(resp, logging_obj=Mock()).results == []
+
+
+def test_transform_search_response_preserves_order():
+    resp = _resp({"results": [_result(title=t) for t in ("first", "second", "third")]})
+    titles = [r.title for r in _config().transform_search_response(resp, logging_obj=Mock()).results]
+    assert titles == ["first", "second", "third"]
\ No newline at end of file