feat(providers): Add LiteLLM provider template support

desktop-shared/src/main/resources/i18n/messages.properties

@@ -979,6 +979,11 @@ provider.template.openrouter.tagline=Access 300+ models — GPT, Claude, Llama,
 provider.template.openrouter.help=💡 To use OpenRouter:\n\n1. Get your free API key at: openrouter.ai/keys\n2. Click Next to browse 300+ available models\n3. Many models have a free tier — no credit card needed
 provider.template.together_ai.tagline=Fast, affordable inference for 100+ open-source models.
 provider.template.together_ai.help=💡 To use Together AI:\n\n1. Get your API key at: api.together.ai/settings/api-keys\n2. Click Next to browse available models,\n   e.g. meta-llama/Llama-3-8b-chat-hf
+provider.template.lite_llm.tagline=A unified proxy that lets you call 100+ LLMs using the OpenAI format.
+provider.search.placeholder=Search providers...
+provider.search.no_results=No providers found
+provider.template.lite_llm.help=💡 To use LiteLLM:\n\n1. Start the LiteLLM proxy: litellm --model ollama/llama3\n2. The default URL is http://localhost:4000/v1\n3. No API key is needed for local setups (set any value if prompted)
+
 
 provider.configure.new.title=Configure {0}
 provider.edit.title=Edit "{0}"

desktop-shared/src/main/resources/i18n/messages_de.properties

@@ -979,6 +979,10 @@ provider.template.openrouter.tagline=Zugriff auf über 300 Modelle — GPT, Clau
 provider.template.openrouter.help=💡 So verwenden Sie OpenRouter:\n\n1. Holen Sie sich Ihren kostenlosen API-Schlüssel unter: openrouter.ai/keys\n2. Klicken Sie auf Weiter, um über 300 verfügbare Modelle zu durchsuchen\n3. Viele Modelle bieten eine kostenlose Stufe — keine Kreditkarte erforderlich
 provider.template.together_ai.tagline=Schnelle, erschwingliche Inferenz für über 100 Open-Source-Modelle.
 provider.template.together_ai.help=💡 So verwenden Sie Together AI:\n\n1. Holen Sie sich Ihren API-Schlüssel unter: api.together.ai/settings/api-keys\n2. Klicken Sie auf Weiter, um verfügbare Modelle zu durchsuchen,\n   z. B. meta-llama/Llama-3-8b-chat-hf
+provider.template.lite_llm.tagline=A unified proxy that lets you call 100+ LLMs using the OpenAI format.
+provider.search.placeholder=Search providers...
+provider.search.no_results=No providers found
+provider.template.lite_llm.help=💡 To use LiteLLM:\n\n1. Start the LiteLLM proxy: litellm --model ollama/llama3\n2. The default URL is http://localhost:4000/v1\n3. No API key is needed for local setups (set any value if prompted)
 
 provider.configure.new.title=Configure {0}
 provider.edit.title=Edit "{0}"

desktop-shared/src/main/resources/i18n/messages_es.properties

@@ -979,6 +979,10 @@ provider.template.openrouter.tagline=Accede a más de 300 modelos — GPT, Claud
 provider.template.openrouter.help=💡 Para usar OpenRouter:\n\n1. Consigue tu clave API gratuita en: openrouter.ai/keys\n2. Haz clic en Siguiente para explorar más de 300 modelos disponibles\n3. Muchos modelos tienen un nivel gratuito — no se requiere tarjeta de crédito
 provider.template.together_ai.tagline=Inferencia rápida y económica para más de 100 modelos de código abierto.
 provider.template.together_ai.help=💡 Para usar Together AI:\n\n1. Consigue tu clave API en: api.together.ai/settings/api-keys\n2. Haz clic en Siguiente para explorar los modelos disponibles,\n   p. ej. meta-llama/Llama-3-8b-chat-hf
+provider.template.lite_llm.tagline=A unified proxy that lets you call 100+ LLMs using the OpenAI format.
+provider.search.placeholder=Search providers...
+provider.search.no_results=No providers found
+provider.template.lite_llm.help=💡 To use LiteLLM:\n\n1. Start the LiteLLM proxy: litellm --model ollama/llama3\n2. The default URL is http://localhost:4000/v1\n3. No API key is needed for local setups (set any value if prompted)
 
 provider.configure.new.title=Configure {0}
 provider.edit.title=Edit "{0}"

desktop-shared/src/main/resources/i18n/messages_fr.properties

@@ -979,6 +979,10 @@ provider.template.openrouter.tagline=Accédez à plus de 300 modèles — GPT, C
 provider.template.openrouter.help=💡 Pour utiliser OpenRouter :\n\n1. Obtenez votre clé API gratuite sur : openrouter.ai/keys\n2. Cliquez sur Suivant pour parcourir plus de 300 modèles disponibles\n3. De nombreux modèles offrent un niveau gratuit — aucune carte de crédit requise
 provider.template.together_ai.tagline=Inférence rapide et abordable pour plus de 100 modèles open source.
 provider.template.together_ai.help=💡 Pour utiliser Together AI :\n\n1. Obtenez votre clé API sur : api.together.ai/settings/api-keys\n2. Cliquez sur Suivant pour parcourir les modèles disponibles,\n   par ex. meta-llama/Llama-3-8b-chat-hf
+provider.template.lite_llm.tagline=A unified proxy that lets you call 100+ LLMs using the OpenAI format.
+provider.search.placeholder=Search providers...
+provider.search.no_results=No providers found
+provider.template.lite_llm.help=💡 To use LiteLLM:\n\n1. Start the LiteLLM proxy: litellm --model ollama/llama3\n2. The default URL is http://localhost:4000/v1\n3. No API key is needed for local setups (set any value if prompted)
 
 provider.configure.new.title=Configure {0}
 provider.edit.title=Edit "{0}"

desktop-shared/src/main/resources/i18n/messages_ja_JP.properties

@@ -979,6 +979,10 @@ provider.template.openrouter.tagline=300 以上のモデルにアクセス — G
 provider.template.openrouter.help=💡 OpenRouter の使い方:\n\n1. openrouter.ai/keys で無料の API キーを取得\n2. 「次へ」をクリックして 300 以上の利用可能なモデルを閲覧\n3. 多くのモデルに無料プランあり — クレジットカード不要
 provider.template.together_ai.tagline=100 以上のオープンソースモデル向けの高速で手頃な推論。
 provider.template.together_ai.help=💡 Together AI の使い方:\n\n1. api.together.ai/settings/api-keys で API キーを取得\n2. 「次へ」をクリックして利用可能なモデルを閲覧\n   （例: meta-llama/Llama-3-8b-chat-hf）
+provider.template.lite_llm.tagline=A unified proxy that lets you call 100+ LLMs using the OpenAI format.
+provider.search.placeholder=Search providers...
+provider.search.no_results=No providers found
+provider.template.lite_llm.help=💡 To use LiteLLM:\n\n1. Start the LiteLLM proxy: litellm --model ollama/llama3\n2. The default URL is http://localhost:4000/v1\n3. No API key is needed for local setups (set any value if prompted)
 
 provider.configure.new.title=Configure {0}
 provider.edit.title=Edit "{0}"

desktop-shared/src/main/resources/i18n/messages_ko_KR.properties

@@ -979,6 +979,10 @@ provider.template.openrouter.tagline=300개 이상의 모델에 액세스 — GP
 provider.template.openrouter.help=💡 OpenRouter 사용 방법:\n\n1. openrouter.ai/keys 에서 무료 API 키를 받으세요\n2. '다음'을 클릭하여 300개 이상의 사용 가능한 모델을 둘러보세요\n3. 많은 모델이 무료 티어를 제공합니다 — 신용카드 불필요
 provider.template.together_ai.tagline=100개 이상의 오픈소스 모델을 위한 빠르고 저렴한 추론.
 provider.template.together_ai.help=💡 Together AI 사용 방법:\n\n1. api.together.ai/settings/api-keys 에서 API 키를 받으세요\n2. '다음'을 클릭하여 사용 가능한 모델을 둘러보세요\n   (예: meta-llama/Llama-3-8b-chat-hf)
+provider.template.lite_llm.tagline=A unified proxy that lets you call 100+ LLMs using the OpenAI format.
+provider.search.placeholder=Search providers...
+provider.search.no_results=No providers found
+provider.template.lite_llm.help=💡 To use LiteLLM:\n\n1. Start the LiteLLM proxy: litellm --model ollama/llama3\n2. The default URL is http://localhost:4000/v1\n3. No API key is needed for local setups (set any value if prompted)
 
 provider.configure.new.title=Configure {0}
 provider.edit.title=Edit "{0}"

desktop-shared/src/main/resources/i18n/messages_pt_BR.properties

@@ -979,6 +979,10 @@ provider.template.openrouter.tagline=Acesse mais de 300 modelos — GPT, Claude,
 provider.template.openrouter.help=💡 Para usar o OpenRouter:\n\n1. Obtenha sua chave API gratuita em: openrouter.ai/keys\n2. Clique em Avançar para explorar mais de 300 modelos disponíveis\n3. Muitos modelos têm um nível gratuito — sem necessidade de cartão de crédito
 provider.template.together_ai.tagline=Inferência rápida e acessível para mais de 100 modelos de código aberto.
 provider.template.together_ai.help=💡 Para usar o Together AI:\n\n1. Obtenha sua chave API em: api.together.ai/settings/api-keys\n2. Clique em Avançar para explorar os modelos disponíveis,\n   ex.: meta-llama/Llama-3-8b-chat-hf
+provider.template.lite_llm.tagline=A unified proxy that lets you call 100+ LLMs using the OpenAI format.
+provider.search.placeholder=Search providers...
+provider.search.no_results=No providers found
+provider.template.lite_llm.help=💡 To use LiteLLM:\n\n1. Start the LiteLLM proxy: litellm --model ollama/llama3\n2. The default URL is http://localhost:4000/v1\n3. No API key is needed for local setups (set any value if prompted)
 
 provider.configure.new.title=Configure {0}
 provider.edit.title=Edit "{0}"

desktop-shared/src/main/resources/i18n/messages_vi_VN.properties

@@ -979,6 +979,10 @@ provider.template.openrouter.tagline=Truy cập hơn 300 mô hình — GPT, Clau
 provider.template.openrouter.help=💡 Để dùng OpenRouter:\n\n1. Lấy API key miễn phí tại: openrouter.ai/keys\n2. Nhấp Tiếp theo để duyệt hơn 300 mô hình có sẵn\n3. Nhiều mô hình có gói miễn phí — không cần thẻ tín dụng
 provider.template.together_ai.tagline=Suy luận nhanh, giá cả phải chăng cho hơn 100 mô hình mã nguồn mở.
 provider.template.together_ai.help=💡 Để dùng Together AI:\n\n1. Lấy API key tại: api.together.ai/settings/api-keys\n2. Nhấp Tiếp theo để duyệt các mô hình có sẵn,\n   ví dụ: meta-llama/Llama-3-8b-chat-hf
+provider.template.lite_llm.tagline=A unified proxy that lets you call 100+ LLMs using the OpenAI format.
+provider.search.placeholder=Search providers...
+provider.search.no_results=No providers found
+provider.template.lite_llm.help=💡 To use LiteLLM:\n\n1. Start the LiteLLM proxy: litellm --model ollama/llama3\n2. The default URL is http://localhost:4000/v1\n3. No API key is needed for local setups (set any value if prompted)
 
 provider.configure.new.title=Configure {0}
 provider.edit.title=Edit "{0}"

desktop-shared/src/main/resources/i18n/messages_zh_CN.properties

@@ -979,6 +979,10 @@ provider.template.openrouter.tagline=通过一个 API 访问 300 多个模型 
 provider.template.openrouter.help=💡 使用 OpenRouter：\n\n1. 在 openrouter.ai/keys 获取免费 API 密钥\n2. 点击"下一步"浏览 300 多个可用模型\n3. 许多模型提供免费额度 — 无需信用卡
 provider.template.together_ai.tagline=为 100 多个开源模型提供快速、经济的推理。
 provider.template.together_ai.help=💡 使用 Together AI：\n\n1. 在 api.together.ai/settings/api-keys 获取 API 密钥\n2. 点击"下一步"浏览可用模型，\n   例如 meta-llama/Llama-3-8b-chat-hf
+provider.template.lite_llm.tagline=A unified proxy that lets you call 100+ LLMs using the OpenAI format.
+provider.search.placeholder=Search providers...
+provider.search.no_results=No providers found
+provider.template.lite_llm.help=💡 To use LiteLLM:\n\n1. Start the LiteLLM proxy: litellm --model ollama/llama3\n2. The default URL is http://localhost:4000/v1\n3. No API key is needed for local setups (set any value if prompted)
 
 provider.configure.new.title=Configure {0}
 provider.edit.title=Edit "{0}"

desktop-shared/src/main/resources/i18n/messages_zh_TW.properties

@@ -979,6 +979,10 @@ provider.template.openrouter.tagline=透過單一 API 存取 300 多個模型 
 provider.template.openrouter.help=💡 若要使用 OpenRouter：\n\n1. 在 openrouter.ai/keys 取得免費 API 金鑰\n2. 按一下「下一步」以瀏覽 300 多個可用模型\n3. 許多模型提供免費方案 — 無需信用卡
 provider.template.together_ai.tagline=為 100 多個開源模型提供快速、實惠的推論。
 provider.template.together_ai.help=💡 若要使用 Together AI：\n\n1. 在 api.together.ai/settings/api-keys 取得 API 金鑰\n2. 按一下「下一步」以瀏覽可用模型，\n   例如 meta-llama/Llama-3-8b-chat-hf
+provider.template.lite_llm.tagline=A unified proxy that lets you call 100+ LLMs using the OpenAI format.
+provider.search.placeholder=Search providers...
+provider.search.no_results=No providers found
+provider.template.lite_llm.help=💡 To use LiteLLM:\n\n1. Start the LiteLLM proxy: litellm --model ollama/llama3\n2. The default URL is http://localhost:4000/v1\n3. No API key is needed for local setups (set any value if prompted)
 
 provider.configure.new.title=Configure {0}
 provider.edit.title=Edit "{0}"

desktop/src/main/kotlin/io/askimo/desktop/settings/ProviderSelectionDialog.kt

@@ -278,47 +278,85 @@ private fun providerTypePickerScreen(viewModel: ProviderWizardViewModel) {
     val mainEntries = remember(entries) { entries.filterNot { it is ProviderEntry.Custom } }
     val hasCustom = remember(entries) { entries.any { it is ProviderEntry.Custom } }
 
+    var searchQuery by remember(viewModel.wizardStep) { mutableStateOf("") }
+    val filteredMainEntries = remember(mainEntries, searchQuery) {
+        if (searchQuery.isBlank()) {
+            mainEntries
+        } else {
+            mainEntries.filter { entry ->
+                val name = when (entry) {
+                    is ProviderEntry.Native -> entry.provider.name
+                    is ProviderEntry.Template -> entry.template.displayName
+                    is ProviderEntry.Custom -> ""
+                }
+                name.contains(searchQuery, ignoreCase = true)
+            }
+        }
+    }
+
     Row(
         modifier = Modifier
             .fillMaxWidth()
             .height(460.dp),
     ) {
-        // ── Left column: scrollable provider list ─────────────────────────────────────────
+        // ── Left column: search field + scrollable provider list ──────────────────────────
         val listState = rememberLazyListState()
-        Box(modifier = Modifier.width(260.dp).fillMaxHeight()) {
-            LazyColumn(state = listState, modifier = Modifier.fillMaxSize()) {
-                items(mainEntries, key = { entry ->
-                    when (entry) {
-                        is ProviderEntry.Native -> "native_${entry.provider.name}"
-                        is ProviderEntry.Template -> "template_${entry.template.name}"
-                        is ProviderEntry.Custom -> "custom"
+        Column(modifier = Modifier.width(260.dp).fillMaxHeight()) {
+            OutlinedTextField(
+                value = searchQuery,
+                onValueChange = { searchQuery = it },
+                modifier = Modifier.fillMaxWidth().padding(horizontal = Spacing.small, vertical = Spacing.extraSmall),
+                placeholder = { Text(stringResource("provider.search.placeholder"), style = AppTextStyles.caption) },
+                singleLine = true,
+                colors = AppComponents.outlinedTextFieldColors(),
+                textStyle = AppTextStyles.caption,
+            )
+            Box(modifier = Modifier.weight(1f)) {
+                LazyColumn(state = listState, modifier = Modifier.fillMaxSize()) {
+                    if (filteredMainEntries.isEmpty() && searchQuery.isNotBlank()) {
+                        item(key = "no_results") {
+                            Text(
+                                text = stringResource("provider.search.no_results"),
+                                style = AppTextStyles.caption,
+                                color = MaterialTheme.colorScheme.onSurfaceVariant,
+                                modifier = Modifier.padding(horizontal = 12.dp, vertical = Spacing.medium),
+                            )
+                        }
+                    } else {
+                        items(filteredMainEntries, key = { entry ->
+                            when (entry) {
+                                is ProviderEntry.Native -> "native_${entry.provider.name}"
+                                is ProviderEntry.Template -> "template_${entry.template.name}"
+                                is ProviderEntry.Custom -> "custom"
+                            }
+                        }) { entry ->
+                            providerPickerEntryRow(
+                                entry = entry,
+                                isSelected = viewModel.selectedEntry == entry,
+                                onClick = { viewModel.selectEntryForPreview(entry) },
+                            )
+                        }
                     }
-                }) { entry ->
-                    providerPickerEntryRow(
-                        entry = entry,
-                        isSelected = viewModel.selectedEntry == entry,
-                        onClick = { viewModel.selectEntryForPreview(entry) },
-                    )
-                }
 
-                if (hasCustom) {
-                    item(key = "custom_divider") {
-                        HorizontalDivider(modifier = Modifier.padding(vertical = Spacing.extraSmall))
-                    }
-                    item(key = "custom_entry") {
-                        providerPickerEntryRow(
-                            entry = ProviderEntry.Custom,
-                            isSelected = viewModel.selectedEntry is ProviderEntry.Custom,
-                            onClick = { viewModel.selectEntryForPreview(ProviderEntry.Custom) },
-                        )
+                    if (hasCustom) {
+                        item(key = "custom_divider") {
+                            HorizontalDivider(modifier = Modifier.padding(vertical = Spacing.extraSmall))
+                        }
+                        item(key = "custom_entry") {
+                            providerPickerEntryRow(
+                                entry = ProviderEntry.Custom,
+                                isSelected = viewModel.selectedEntry is ProviderEntry.Custom,
+                                onClick = { viewModel.selectEntryForPreview(ProviderEntry.Custom) },
+                            )
+                        }
                     }
                 }
+                VerticalScrollbar(
+                    adapter = rememberScrollbarAdapter(listState),
+                    modifier = Modifier.align(Alignment.CenterEnd).fillMaxHeight().padding(end = 2.dp),
+                    style = AppComponents.scrollbarStyle(),
+                )
             }
-            VerticalScrollbar(
-                adapter = rememberScrollbarAdapter(listState),
-                modifier = Modifier.align(Alignment.CenterEnd).fillMaxHeight().padding(end = 2.dp),
-                style = AppComponents.scrollbarStyle(),
-            )
         }
 
         // ── Vertical divider ──────────────────────────────────────────────────────────────

gradle.properties

@@ -1,6 +1,6 @@
 # Project metadata
 projectGroup=io.askimo
-projectVersion=1.4.16
+projectVersion=1.4.17
 jvmVersion=25
 
 # About information

shared/src/main/kotlin/io/askimo/core/providers/openaicompatible/OpenAiCompatibleTemplate.kt

@@ -120,4 +120,14 @@ enum class OpenAiCompatibleTemplate(
         httpVersion = HttpVersion.HTTP_1_1,
         helpTextKey = "provider.template.together_ai.help",
     ),
+
+    LITE_LLM(
+        displayName = "LiteLLM",
+        initials = "LL",
+        taglineKey = "provider.template.lite_llm.tagline",
+        baseUrl = "http://localhost:4000/v1",
+        apiKeyRequired = false,
+        apiKeyUrl = "",
+        helpTextKey = "provider.template.lite_llm.help",
+    ),
 }