diff --git a/_assets/plugin_preview.png b/_assets/plugin_preview.png
index a084f47..8f66afa 100644
Binary files a/_assets/plugin_preview.png and b/_assets/plugin_preview.png differ
diff --git a/_assets/qiniu_ai.png b/_assets/qiniu_ai.png
deleted file mode 100644
index bd06985..0000000
Binary files a/_assets/qiniu_ai.png and /dev/null differ
diff --git a/manifest.yaml b/manifest.yaml
index ace90de..addcf1e 100644
--- a/manifest.yaml
+++ b/manifest.yaml
@@ -35,4 +35,4 @@ resource:
     tool:
       enabled: false
 type: plugin
-version: 0.0.2
+version: 0.0.3
diff --git a/models/llm/glm45-air.yaml b/models/llm/glm45-air.yaml
new file mode 100644
index 0000000..129c673
--- /dev/null
+++ b/models/llm/glm45-air.yaml
@@ -0,0 +1,71 @@
+model: glm-4.5-air
+label:
+  zh_Hans: glm-4.5-air
+  en_US: glm-4.5-air
+model_type: llm
+features:
+  - agent-thought
+  - multi-tool-call
+  - stream-tool-call
+model_properties:
+  mode: chat
+  context_size: 128000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+  - name: max_tokens
+    use_template: max_tokens
+    type: int
+    default: 512
+    min: 1
+    max: 8192
+    help:
+      zh_Hans: 指定生成结果长度的上限。如果生成结果截断，可以调大该参数。
+      en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter.
+  - name: top_p
+    use_template: top_p
+  - name: top_k
+    label:
+      zh_Hans: 取样数量
+      en_US: Top k
+    type: int
+    help:
+      zh_Hans: 仅从每个后续标记的前 K 个选项中采样。
+      en_US: Only sample from the top K options for each subsequent token.
+    required: false
+  - name: frequency_penalty
+    use_template: frequency_penalty
+  - name: response_format
+    label:
+      zh_Hans: 回复格式
+      en_US: Response Format
+    type: string
+    help:
+      zh_Hans: 指定模型必须输出的格式
+      en_US: specifying the format that the model must output
+    required: false
+    options:
+      - text
+      - json_object
+  - name: enable_thinking
+    required: false
+    type: boolean
+    default: true
+    label:
+      zh_Hans: 思考模式
+      en_US: Thinking mode
+    help:
+      zh_Hans: 是否开启思考模式。
+      en_US: Whether to enable thinking mode.
+  - name: thinking_budget
+    required: false
+    type: int
+    default: 512
+    min: 1
+    max: 8192
+    label:
+      zh_Hans: 思考长度限制
+      en_US: Thinking budget
+    help:
+      zh_Hans: 思考过程的最大长度，只在思考模式为true时生效。
+      en_US: The maximum length of the thinking process, only effective when thinking mode is true.
diff --git a/models/llm/glm45.yaml b/models/llm/glm45.yaml
new file mode 100644
index 0000000..2a0ee74
--- /dev/null
+++ b/models/llm/glm45.yaml
@@ -0,0 +1,71 @@
+model: glm-4.5
+label:
+  zh_Hans: glm-4.5
+  en_US: glm-4.5
+model_type: llm
+features:
+  - agent-thought
+  - multi-tool-call
+  - stream-tool-call
+model_properties:
+  mode: chat
+  context_size: 128000
+parameter_rules:
+  - name: temperature
+    use_template: temperature
+  - name: max_tokens
+    use_template: max_tokens
+    type: int
+    default: 512
+    min: 1
+    max: 8192
+    help:
+      zh_Hans: 指定生成结果长度的上限。如果生成结果截断，可以调大该参数。
+      en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter.
+  - name: top_p
+    use_template: top_p
+  - name: top_k
+    label:
+      zh_Hans: 取样数量
+      en_US: Top k
+    type: int
+    help:
+      zh_Hans: 仅从每个后续标记的前 K 个选项中采样。
+      en_US: Only sample from the top K options for each subsequent token.
+    required: false
+  - name: frequency_penalty
+    use_template: frequency_penalty
+  - name: response_format
+    label:
+      zh_Hans: 回复格式
+      en_US: Response Format
+    type: string
+    help:
+      zh_Hans: 指定模型必须输出的格式
+      en_US: specifying the format that the model must output
+    required: false
+    options:
+      - text
+      - json_object
+  - name: enable_thinking
+    required: false
+    type: boolean
+    default: true
+    label:
+      zh_Hans: 思考模式
+      en_US: Thinking mode
+    help:
+      zh_Hans: 是否开启思考模式。
+      en_US: Whether to enable thinking mode.
+  - name: thinking_budget
+    required: false
+    type: int
+    default: 512
+    min: 1
+    max: 8192
+    label:
+      zh_Hans: 思考长度限制
+      en_US: Thinking budget
+    help:
+      zh_Hans: 思考过程的最大长度，只在思考模式为true时生效。
+      en_US: The maximum length of the thinking process, only effective when thinking mode is true.
diff --git a/provider/qiniu.yaml b/provider/qiniu.yaml
index 2174388..c2daf93 100644
--- a/provider/qiniu.yaml
+++ b/provider/qiniu.yaml
@@ -3,8 +3,8 @@ label:
   en_US: Qiniu Cloud
   zh_Hans: 七牛云
 description:
-  en_US: Official Qiniu Dify plugin providing AI inference services, supporting models such as deepseek-r1, deepseek-v3, and more.
-  zh_Hans: 七牛云官方 Dify 插件，提供 AI 推理服务，支持例如 deepseek-r1、deepseek-v3 等模型。
+  en_US: Official Qiniu Dify plugin providing AI inference services, supporting models such as glm 4.5, deepseek-r1, deepseek-v3, and more.
+  zh_Hans: 七牛云官方 Dify 插件，提供 AI 推理服务，支持例如 glm 4.5、deepseek-r1、deepseek-v3 等模型。
 icon_large:
   en_US: icon_l_en.svg
 icon_small:
diff --git a/requirements.txt b/requirements.txt
index 81049e8..b267cde 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1 +1 @@
-dify_plugin>=0.3.0,<0.4.0
+dify_plugin>=0.3.0,<0.5.0