diff --git a/plugins/model-providers/upstage/__init__.py b/plugins/model-providers/upstage/__init__.py
index 03e954e3e4ea..982e741f9e86 100644
--- a/plugins/model-providers/upstage/__init__.py
+++ b/plugins/model-providers/upstage/__init__.py
@@ -6,10 +6,12 @@ from providers import register_provider
 from providers.base import ProviderProfile
 
 
-# Model-name markers for Solar families that accept ``reasoning_effort``.
-# Substring match (not startswith) so dated variants like
-# ``solar-pro3-250127`` and ``solar-open-...`` are covered too.
-_REASONING_MODEL_MARKERS = ("solar-pro", "solar-open")
+# Model-name markers for Solar families that do NOT accept ``reasoning_effort``.
+# Deny-list on purpose: newly released Solar models are assumed
+# reasoning-capable by default, so only the known non-reasoning families are
+# listed here. Substring match (not startswith) so dated variants like
+# ``solar-mini-250127`` are covered too.
+_NON_REASONING_MODEL_MARKERS = ("solar-mini", "syn-pro")
 
 # When the user hasn't picked a reasoning effort, Hermes passes
 # reasoning_config=None. Solar's own server default is "minimal" (reasoning
@@ -22,15 +24,20 @@ _DEFAULT_REASONING_EFFORT = "medium"
 
 
 def _model_supports_reasoning(model: str | None) -> bool:
-    """Solar reasoning-capable models.
+    """Solar reasoning-capable models — True unless the model is deny-listed.
 
     The Solar Pro family (``solar-pro``, ``solar-pro2``, ``solar-pro3`` and
     dated variants like ``solar-pro3-250127``) and the Solar Open family
-    (``solar-open*``) accept ``reasoning_effort``. ``solar-mini`` / ``syn-pro``
-    ignore the parameter, so we don't send it.
+    (``solar-open*``) accept ``reasoning_effort``; only ``solar-mini`` /
+    ``syn-pro`` ignore the parameter, so we deny-list those and treat every
+    other (incl. future) Solar model as reasoning-capable.
+
+    ``None``/empty model → True: the provider default (``fallback_models[0]``,
+    the ``solar-pro`` rolling alias) is reasoning-capable, so an unset model
+    gets the same default-on behaviour.
     """
     m = (model or "").strip().lower()
-    return any(marker in m for marker in _REASONING_MODEL_MARKERS)
+    return not any(marker in m for marker in _NON_REASONING_MODEL_MARKERS)
 
 
 class UpstageProfile(ProviderProfile):
@@ -54,7 +61,8 @@ class UpstageProfile(ProviderProfile):
     ) -> tuple[dict[str, Any], dict[str, Any]]:
         top_level: dict[str, Any] = {}
 
-        # solar-mini / syn-pro ignore reasoning_effort — send nothing.
+        # solar-mini / syn-pro (the deny-list) ignore reasoning_effort — send
+        # nothing. Everything else, including future Solar models, gets it.
         if not _model_supports_reasoning(model):
             return {}, top_level
 
diff --git a/tests/plugins/model_providers/test_upstage_profile.py b/tests/plugins/model_providers/test_upstage_profile.py
index 591408d71b76..e78c2326ee11 100644
--- a/tests/plugins/model_providers/test_upstage_profile.py
+++ b/tests/plugins/model_providers/test_upstage_profile.py
@@ -113,9 +113,9 @@ class TestUpstageReasoning:
         _, top_level = upstage_profile.build_api_kwargs_extras(model=model)
         assert top_level == {"reasoning_effort": "medium"}
 
-    @pytest.mark.parametrize("model", ["solar-mini", "syn-pro"])
-    def test_no_config_non_pro_still_omits(self, upstage_profile, model):
-        # Default-on must not leak to non-reasoning models.
+    @pytest.mark.parametrize("model", ["solar-mini", "solar-mini-202610", "syn-pro"])
+    def test_no_config_deny_listed_still_omits(self, upstage_profile, model):
+        # Default-on must not leak to the deny-listed non-reasoning models.
         _, top_level = upstage_profile.build_api_kwargs_extras(model=model)
         assert top_level == {}
 
@@ -137,15 +137,36 @@ class TestUpstageReasoning:
         )
         assert top_level == {"reasoning_effort": "high"}
 
-    @pytest.mark.parametrize("model", ["solar-mini", "syn-pro"])
-    def test_non_pro_models_never_send_reasoning(self, upstage_profile, model):
-        # solar-mini / syn-pro ignore reasoning_effort, so never send it.
+    @pytest.mark.parametrize("model", ["solar-mini", "solar-mini-202610", "syn-pro"])
+    def test_deny_listed_models_never_send_reasoning(self, upstage_profile, model):
+        # solar-mini / syn-pro ignore reasoning_effort, so never send it —
+        # even when the user explicitly enables reasoning.
         extra_body, top_level = upstage_profile.build_api_kwargs_extras(
             reasoning_config={"enabled": True, "effort": "high"}, model=model
         )
         assert extra_body == {}
         assert top_level == {}
 
+    @pytest.mark.parametrize("model", ["solar-future", "solar-future-260601"])
+    def test_unknown_future_models_default_to_reasoning(self, upstage_profile, model):
+        # Deny-list semantics: a future Solar model we've never heard of is
+        # assumed reasoning-capable, so reasoning_effort is sent instead of
+        # being silently dropped (the old allow-list failure mode).
+        _, top_level = upstage_profile.build_api_kwargs_extras(
+            reasoning_config={"enabled": True, "effort": "high"}, model=model
+        )
+        assert top_level == {"reasoning_effort": "high"}
+
+        # And the unset-config default-on path applies to it too.
+        _, top_level = upstage_profile.build_api_kwargs_extras(model=model)
+        assert top_level == {"reasoning_effort": "medium"}
+
+    def test_none_model_defaults_to_reasoning(self, upstage_profile):
+        # No model in context → treated as reasoning-capable, consistent with
+        # the provider default (fallback_models[0] == "solar-pro").
+        _, top_level = upstage_profile.build_api_kwargs_extras(model=None)
+        assert top_level == {"reasoning_effort": "medium"}
+
 
 def upstage_profile_singleton():
     import providers