switch to using gpt-5-mini for free proxy (#2297)

Author: sawka

aiprompts/config-system.md

@@ -41,6 +41,7 @@ waveterm/
 ```
 
 **Key Files:**
+
 - **[`pkg/wconfig/settingsconfig.go`](pkg/wconfig/settingsconfig.go)** - Defines the `SettingsType` struct with all configuration fields
 - **[`schema/settings.json`](schema/settings.json)** - JSON Schema for validation and type checking
 - **[`pkg/wconfig/defaultconfig/settings.json`](pkg/wconfig/defaultconfig/settings.json)** - Default values for all settings
@@ -65,10 +66,12 @@ Wave Terminal supports block-level configuration overrides through the metadata
 3. **Block Metadata** - Override settings for individual blocks
 
 **Key Files for Block Overrides:**
+
 - **[`pkg/waveobj/wtypemeta.go`](pkg/waveobj/wtypemeta.go)** - Defines the `MetaTSType` struct for block-level metadata
 - Block metadata fields should match the corresponding settings fields for consistency
 
 **Frontend Usage:**
+
 ```typescript
 // Use getOverrideConfigAtom for hierarchical config resolution
 const settingValue = useAtomValue(getOverrideConfigAtom(blockId, "namespace:setting"));
@@ -77,6 +80,7 @@ const settingValue = useAtomValue(getOverrideConfigAtom(blockId, "namespace:sett
 ```
 
 **Setting Block Metadata:**
+
 ```bash
 # Set for current block
 wsh setmeta namespace:setting=value
@@ -143,6 +147,7 @@ type MetaTSType struct {
 ```
 
 **Block Metadata Guidelines:**
+
 - Use pointer types (`*string`, `*bool`, `*int`, `*float64`) for optional overrides
 - JSON tags should exactly match the corresponding settings field
 - This enables the hierarchical config system: block metadata → connection config → global settings
@@ -197,7 +202,7 @@ If your setting should have a default value, add it to [`pkg/wconfig/defaultconf
 ```json
 {
   "ai:preset": "ai@global",
-  "ai:model": "gpt-4o-mini",
+  "ai:model": "gpt-5-mini",
   // ... existing defaults ...
 
   "mynew:setting": "default value",
@@ -239,6 +244,7 @@ task generate
 ```
 
 Or run them individually:
+
 ```bash
 # Regenerate JSON schema
 task build:schema

docs/docs/config.mdx

@@ -100,8 +100,8 @@ For reference, this is the current default configuration (v0.11.5):
 ```json
 {
     "ai:preset": "ai@global",
-    "ai:model": "gpt-4o-mini",
-    "ai:maxtokens": 2048,
+    "ai:model": "gpt-5-mini",
+    "ai:maxtokens": 4000,
     "ai:timeoutms": 60000,
     "app:defaultnewblock": "term",
     "autoupdate:enabled": true,

frontend/app/view/waveai/waveai.tsx

@@ -223,7 +223,7 @@ export class WaveAiModel implements ViewModel {
                         viewTextChildren.push({
                             elemtype: "iconbutton",
                             icon: "cloud",
-                            title: "Using Wave's AI Proxy (gpt-4o-mini)",
+                            title: "Using Wave's AI Proxy (gpt-5-mini)",
                             noAction: true,
                         });
                     } else {

pkg/wconfig/defaultconfig/presets/ai.json

@@ -5,16 +5,16 @@
         "ai:*": true
     },
     "ai@wave": {
-        "display:name": "Wave Proxy - gpt-4o-mini",
+        "display:name": "Wave Proxy - gpt-5-mini",
         "display:order": 0,
         "ai:*": true,
         "ai:apitype": "",
         "ai:baseurl": "",
         "ai:apitoken": "",
         "ai:name": "",
         "ai:orgid": "",
-        "ai:model": "gpt-4o-mini",
-        "ai:maxtokens": 2048,
+        "ai:model": "gpt-5-mini",
+        "ai:maxtokens": 4000,
         "ai:timeoutms": 60000
     }
 }

pkg/wconfig/defaultconfig/settings.json

@@ -1,7 +1,7 @@
 {
     "ai:preset": "ai@global",
-    "ai:model": "gpt-4o-mini",
-    "ai:maxtokens": 2048,
+    "ai:model": "gpt-5-mini",
+    "ai:maxtokens": 4000,
     "ai:timeoutms": 60000,
     "app:defaultnewblock": "term",
     "autoupdate:enabled": true,