From fbeffe766bbfd5b9b0ccba52cabc0ab5bb1434bb Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E6=88=91=E4=BB=AC=E5=9C=A8?= <1244791533@qq.com>
Date: Wed, 16 Sep 2026 16:16:32 +0800
Subject: [PATCH 1/2] Add volcengine GLM-5.3-Flash (glm-5-3-flash-260828)

---
 .../volcengine/models/glm-5-3-flash-260828.toml      | 12 ++++++++++++
 1 file changed, 12 insertions(+)
 create mode 100644 providers/volcengine/models/glm-5-3-flash-260828.toml

diff --git a/providers/volcengine/models/glm-5-3-flash-260828.toml b/providers/volcengine/models/glm-5-3-flash-260828.toml
new file mode 100644
index 00000000000..44cd5cf1527
--- /dev/null
+++ b/providers/volcengine/models/glm-5-3-flash-260828.toml
@@ -0,0 +1,12 @@
+# Model ID verified against POST /api/v3/chat/completions (2026-09-16): `glm-5-3-flash-260828`
+base_model = "zhipuai/glm-5.3-flash"
+
+[[reasoning_options]]
+type = "toggle"
+
+[[reasoning_options]]
+type = "effort"
+values = ["minimal", "low", "medium", "high"]
+
+[interleaved]
+field = "reasoning_content"

From b10b451f51d3dac28a7e2fa20f18616ad3669477 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E6=88=91=E4=BB=AC=E5=9C=A8?= <1244791533@qq.com>
Date: Wed, 16 Sep 2026 16:25:18 +0800
Subject: [PATCH 2/2] Fix reasoning options and add cost per review

---
 .../volcengine/models/glm-5-3-flash-260828.toml    | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

diff --git a/providers/volcengine/models/glm-5-3-flash-260828.toml b/providers/volcengine/models/glm-5-3-flash-260828.toml
index 44cd5cf1527..98cb87ee9ba 100644
--- a/providers/volcengine/models/glm-5-3-flash-260828.toml
+++ b/providers/volcengine/models/glm-5-3-flash-260828.toml
@@ -1,12 +1,18 @@
 # Model ID verified against POST /api/v3/chat/completions (2026-09-16): `glm-5-3-flash-260828`
+# CNY→USD rate: 6.737012, 2026-08-26, source: https://open.er-api.com/v6/latest/USD
+# CNY list price source: https://docs.volcengine.com/docs/82379/1449737 (深度思考) + 智谱官方价目表
+# GLM-5.3-Flash always reasons (thinking cannot be disabled on this model ID)
+# Effort: reasoning_effort = low|high|max
 base_model = "zhipuai/glm-5.3-flash"
 
-[[reasoning_options]]
-type = "toggle"
-
 [[reasoning_options]]
 type = "effort"
-values = ["minimal", "low", "medium", "high"]
+values = ["low", "high", "max"]
 
 [interleaved]
 field = "reasoning_content"
+
+[cost]
+input = 0.11875
+output = 0.41563
+cache_read = 0.03414