1# Agent Settings
2
3Learn about all the settings you can customize in Zed's Agent Panel.
4
5## Model Settings {#model-settings}
6
7### Default Model {#default-model}
8
9If you're using [Zed's hosted LLM service](./subscription.md), it sets `claude-sonnet-4` as the default model for agentic work (agent panel, inline assistant) and `gpt-5-nano` as the default "fast" model (thread summarization, git commit messages). If you're not subscribed or want to change these defaults, you can manually edit the `default_model` object in your settings:
10
11```json [settings]
12{
13 "agent": {
14 "default_model": {
15 "provider": "openai",
16 "model": "gpt-4o"
17 }
18 }
19}
20```
21
22### Feature-specific Models {#feature-specific-models}
23
24You can assign distinct and specific models for the following AI-powered features:
25
26- Thread summary model: Used for generating thread summaries
27- Inline assistant model: Used for the inline assistant feature
28- Commit message model: Used for generating Git commit messages
29
30```json [settings]
31{
32 "agent": {
33 "default_model": {
34 "provider": "zed.dev",
35 "model": "claude-sonnet-4"
36 },
37 "inline_assistant_model": {
38 "provider": "anthropic",
39 "model": "claude-3-5-sonnet"
40 },
41 "commit_message_model": {
42 "provider": "openai",
43 "model": "gpt-4o-mini"
44 },
45 "thread_summary_model": {
46 "provider": "google",
47 "model": "gemini-2.0-flash"
48 }
49 }
50}
51```
52
53> If a custom model isn't set for one of these features, they automatically fall back to using the default model.
54
55### Alternative Models for Inline Assists {#alternative-assists}
56
57The Inline Assist feature in particular has the capacity to perform multiple generations in parallel using different models.
58That is possible by assigning more than one model to it, taking the configuration shown above one step further.
59
60When configured, the inline assist UI will surface controls to cycle between the outputs generated by each model.
61
62The models you specify here are always used in _addition_ to your [default model](#default-model).
63
64For example, the following configuration will generate two outputs for every assist.
65One with Claude Sonnet 4 (the default model), and one with GPT-5-mini.
66
67```json [settings]
68{
69 "agent": {
70 "default_model": {
71 "provider": "zed.dev",
72 "model": "claude-sonnet-4"
73 },
74 "inline_alternatives": [
75 {
76 "provider": "zed.dev",
77 "model": "gpt-4-mini"
78 }
79 ]
80 }
81}
82```
83
84### Model Temperature
85
86Specify a custom temperature for a provider and/or model:
87
88```json [settings]
89"model_parameters": [
90 // To set parameters for all requests to OpenAI models:
91 {
92 "provider": "openai",
93 "temperature": 0.5
94 },
95 // To set parameters for all requests in general:
96 {
97 "temperature": 0
98 },
99 // To set parameters for a specific provider and model:
100 {
101 "provider": "zed.dev",
102 "model": "claude-sonnet-4",
103 "temperature": 1.0
104 }
105],
106```
107
108## Agent Panel Settings {#agent-panel-settings}
109
110Note that some of these settings are also surfaced in the Agent Panel's settings UI, which you can access either via the `agent: open settings` action or by the dropdown menu on the top-right corner of the panel.
111
112### Default View
113
114Use the `default_view` setting to change the default view of the Agent Panel.
115You can choose between `thread` (the default) and `text_thread`:
116
117```json [settings]
118{
119 "agent": {
120 "default_view": "text_thread"
121 }
122}
123```
124
125### Font Size
126
127Use the `agent_font_size` setting to change the font size of rendered agent responses in the panel.
128
129```json [settings]
130{
131 "agent": {
132 "agent_font_size": 18
133 }
134}
135```
136
137> Editors in the Agent Panel—whether that is the main message textarea or previous messages—use monospace fonts and therefore, are controlled by the `buffer_font_size` setting, which is defined globally in your `settings.json`.
138
139### Auto-run Commands
140
141Control whether to allow the agent to run commands without asking you for permission.
142The default value is `false`.
143
144```json [settings]
145{
146 "agent": {
147 "always_allow_tool_actions": true
148 }
149}
150```
151
152### Single-file Review
153
154Control whether to display review actions (accept & reject) in single buffers after the agent is done performing edits.
155The default value is `false`.
156
157```json [settings]
158{
159 "agent": {
160 "single_file_review": true
161 }
162}
163```
164
165When set to false, these controls are only available in the multibuffer review tab.
166
167### Sound Notification
168
169Control whether to hear a notification sound when the agent is done generating changes or needs your input.
170The default value is `false`.
171
172```json [settings]
173{
174 "agent": {
175 "play_sound_when_agent_done": true
176 }
177}
178```
179
180### Message Editor Size
181
182Use the `message_editor_min_lines` setting to control minimum number of lines of height the agent message editor should have.
183It is set to `4` by default, and the max number of lines is always double of the minimum.
184
185```json [settings]
186{
187 "agent": {
188 "message_editor_min_lines": 4
189 }
190}
191```
192
193### Modifier to Send
194
195Make a modifier (`cmd` on macOS, `ctrl` on Linux) required to send messages.
196This is encouraged for more thoughtful prompt crafting.
197The default value is `false`.
198
199```json [settings]
200{
201 "agent": {
202 "use_modifier_to_send": true
203 }
204}
205```
206
207### Edit Card
208
209Use the `expand_edit_card` setting to control whether edit cards show the full diff in the Agent Panel.
210It is set to `true` by default, but if set to false, the card's height is capped to a certain number of lines, requiring a click to be expanded.
211
212```json [settings]
213{
214 "agent": {
215 "expand_edit_card": false
216 }
217}
218```
219
220### Terminal Card
221
222Use the `expand_terminal_card` setting to control whether terminal cards show the command output in the Agent Panel.
223It is set to `true` by default, but if set to false, the card will be fully collapsed even while the command is running, requiring a click to be expanded.
224
225```json [settings]
226{
227 "agent": {
228 "expand_terminal_card": false
229 }
230}
231```
232
233### Feedback Controls
234
235Control whether to display the thumbs up/down buttons at the bottom of each agent response, allowing to give Zed feedback about the agent's performance.
236The default value is `true`.
237
238```json [settings]
239{
240 "agent": {
241 "enable_feedback": false
242 }
243}
244```