Add NVIDIA NIM API provider support

src/background/index.mjs

@@ -30,6 +30,7 @@ import {
   isUsingOpenRouterApiModel,
   isUsingAimlApiModel,
   isUsingDeepSeekApiModel,
+  isUsingNvidiaNimApiModel,
   isUsingGoogleApiModel,
 } from '../config/index.mjs'
 import '../_locales/i18n'
@@ -398,6 +399,7 @@ function isUsingOpenAICompatibleApiSession(session) {
     isUsingMoonshotApiModel(session) ||
     isUsingChatGLMApiModel(session) ||
     isUsingDeepSeekApiModel(session) ||
+    isUsingNvidiaNimApiModel(session) ||
     isUsingOllamaApiModel(session) ||
     isUsingOpenRouterApiModel(session) ||
     isUsingAimlApiModel(session) ||

src/config/index.mjs

@@ -129,6 +129,15 @@ export const deepSeekApiModelKeys = [
   'deepseek_v4_flash',
   'deepseek_v4_pro',
 ]
+export const nvidiaNimApiModelKeys = [
+  'nvidiaNim_nemotron_3_super',
+  'nvidiaNim_nemotron_3_ultra',
+  'nvidiaNim_deepseek_v4_flash',
+  'nvidiaNim_deepseek_v4_pro',
+  'nvidiaNim_glm_5_2',
+  'nvidiaNim_inkling',
+  'nvidiaNim_minimax_m3',
+]
 export const openRouterApiModelKeys = [
   'openRouter_auto',
   'openRouter_free',
@@ -249,6 +258,10 @@ export const ModelGroups = {
     value: githubThirdPartyApiModelKeys,
     desc: 'Github Third Party Waylaidwanderer (API)',
   },
+  nvidiaNimApiModelKeys: {
+    value: nvidiaNimApiModelKeys,
+    desc: 'NVIDIA NIM (API)',
+  },
   deepSeekApiModelKeys: {
     value: deepSeekApiModelKeys,
     desc: 'DeepSeek (API)',
@@ -415,6 +428,35 @@ export const Models = {
     desc: 'Kimi.Moonshot (128k)',
   },
 
+  nvidiaNim_nemotron_3_super: {
+    value: 'nvidia/nemotron-3-super-120b-a12b',
+    desc: 'NVIDIA NIM (Nemotron-3-Super)',
+  },
+  nvidiaNim_nemotron_3_ultra: {
+    value: 'nvidia/nemotron-3-ultra-550b-a55b',
+    desc: 'NVIDIA NIM (Nemotron-3-Ultra)',
+  },
+  nvidiaNim_deepseek_v4_flash: {
+    value: 'deepseek-ai/deepseek-v4-flash',
+    desc: 'NVIDIA NIM (DeepSeek V4 Flash)',
+  },
+  nvidiaNim_deepseek_v4_pro: {
+    value: 'deepseek-ai/deepseek-v4-pro',
+    desc: 'NVIDIA NIM (DeepSeek V4 Pro)',
+  },
+  nvidiaNim_glm_5_2: {
+    value: 'z-ai/glm-5.2',
+    desc: 'NVIDIA NIM (GLM-5.2)',
+  },
+  nvidiaNim_inkling: {
+    value: 'thinkingmachines/inkling',
+    desc: 'NVIDIA NIM (Inkling)',
+  },
+  nvidiaNim_minimax_m3: {
+    value: 'minimaxai/minimax-m3',
+    desc: 'NVIDIA NIM (MiniMax M3)',
+  },
+
   deepseek_chat: {
     value: 'deepseek-chat',
     desc: 'DeepSeek (Chat)',
@@ -692,6 +734,7 @@ export const defaultConfig = {
   chatglmApiKey: '',
   moonshotApiKey: '',
   deepSeekApiKey: '',
+  nvidiaNimApiKey: '',
 
   customApiKey: '',
 
@@ -752,6 +795,7 @@ export const defaultConfig = {
     'googleGemini3_5Flash',
     'openRouter_auto',
     'openRouter_free',
+    'nvidiaNim_nemotron_3_super',
   ],
   customApiModes: [
     {
@@ -899,6 +943,10 @@ export function isUsingDeepSeekApiModel(configOrSession) {
   return isInApiModeGroup(deepSeekApiModelKeys, configOrSession)
 }
 
+export function isUsingNvidiaNimApiModel(configOrSession) {
+  return isInApiModeGroup(nvidiaNimApiModelKeys, configOrSession)
+}
+
 export function isUsingOpenRouterApiModel(configOrSession) {
   return isInApiModeGroup(openRouterApiModelKeys, configOrSession)
 }

src/config/openai-provider-mappings.mjs

@@ -1,6 +1,7 @@
 export const LEGACY_API_KEY_FIELD_BY_PROVIDER_ID = {
   openai: 'apiKey',
   deepseek: 'deepSeekApiKey',
+  'nvidia-nim': 'nvidiaNimApiKey',
   moonshot: 'moonshotApiKey',
   openrouter: 'openRouterApiKey',
   aiml: 'aimlApiKey',
@@ -22,6 +23,7 @@ export const OPENAI_COMPATIBLE_GROUP_TO_PROVIDER_ID = {
   gptApiModelKeys: 'openai',
   moonshotApiModelKeys: 'moonshot',
   deepSeekApiModelKeys: 'deepseek',
+  nvidiaNimApiModelKeys: 'nvidia-nim',
   openRouterApiModelKeys: 'openrouter',
   aimlModelKeys: 'aiml',
   aimlApiModelKeys: 'aiml',

src/popup/sections/GeneralPart.jsx

@@ -58,6 +58,8 @@ function isUsingSpecialCustomModel(configOrSession) {
 
 function getProviderApiKeySetupUrl(providerId) {
   switch (String(providerId || '').trim()) {
+    case 'nvidia-nim':
+      return 'https://build.nvidia.com/settings/keys'
     case 'openai':
       return 'https://platform.openai.com/account/api-keys'
     case 'moonshot':

src/services/apis/provider-registry.mjs

@@ -26,6 +26,14 @@ const BUILTIN_PROVIDER_TEMPLATE = [
     builtin: true,
     enabled: true,
   },
+  {
+    id: 'nvidia-nim',
+    name: 'NVIDIA NIM',
+    baseUrl: 'https://integrate.api.nvidia.com',
+    chatCompletionsPath: '/v1/chat/completions',
+    builtin: true,
+    enabled: true,
+  },
   {
     id: 'moonshot',
     name: 'Kimi.Moonshot',
@@ -110,6 +118,7 @@ function resolveProviderIdFromLegacyModelName(modelName) {
     return 'openai'
   }
   if (preset.startsWith('deepseek_') || preset === 'deepSeekApiModelKeys') return 'deepseek'
+  if (preset.startsWith('nvidiaNim_') || preset === 'nvidiaNimApiModelKeys') return 'nvidia-nim'
   if (preset.startsWith('moonshot_') || preset === 'moonshotApiModelKeys') return 'moonshot'
   if (preset.startsWith('openRouter_') || preset === 'openRouterApiModelKeys') return 'openrouter'
   if (preset.startsWith('aiml_') || preset === 'aimlModelKeys' || preset === 'aimlApiModelKeys') {

tests/unit/config/config-predicates.test.mjs

@@ -8,6 +8,7 @@ import {
   gptApiModelKeys,
   claudeApiModelKeys,
   openRouterApiModelKeys,
+  nvidiaNimApiModelKeys,
   aimlApiModelKeys,
   isUsingAimlApiModel,
   isUsingAzureOpenAiApiModel,
@@ -19,6 +20,7 @@ import {
   isUsingCustomModel,
   isUsingCustomNameOnlyModel,
   isUsingDeepSeekApiModel,
+  isUsingNvidiaNimApiModel,
   isUsingGeminiWebModel,
   isUsingGithubThirdPartyApiModel,
   isUsingMoonshotApiModel,
@@ -229,6 +231,13 @@ test('isUsingDeepSeekApiModel detects DeepSeek models', () => {
   assert.equal(isUsingDeepSeekApiModel({ modelName: 'chatgptApi4oMini' }), false)
 })
 
+test('isUsingNvidiaNimApiModel accepts exported NVIDIA NIM API model keys', () => {
+  for (const modelName of nvidiaNimApiModelKeys) {
+    assert.equal(isUsingNvidiaNimApiModel({ modelName }), true)
+  }
+  assert.equal(isUsingNvidiaNimApiModel({ modelName: 'chatgptApi4oMini' }), false)
+})
+
 test('isUsingOpenRouterApiModel matches representative OpenRouter API keys', () => {
   for (const modelName of representativeOpenRouterApiModelNames) {
     assert.equal(isUsingOpenRouterApiModel({ modelName }), true)

tests/unit/services/apis/provider-registry.test.mjs

@@ -46,6 +46,17 @@ test('resolveProviderIdForSession resolves legacy Google preset keys', () => {
   assert.equal(resolveProviderIdForSession({ modelName: 'googleApiModelKeys-custom' }), 'google')
 })
 
+test('resolveProviderIdForSession resolves legacy NVIDIA NIM preset keys', () => {
+  assert.equal(
+    resolveProviderIdForSession({ modelName: 'nvidiaNim_nemotron_3_super' }),
+    'nvidia-nim',
+  )
+  assert.equal(
+    resolveProviderIdForSession({ modelName: 'nvidiaNimApiModelKeys-custom' }),
+    'nvidia-nim',
+  )
+})
+
 test('resolveProviderIdForSession does not treat raw Gemini model IDs as legacy presets', () => {
   assert.equal(resolveProviderIdForSession({ modelName: 'gemini-2.5-flash' }), null)
 })

tests/unit/services/apis/thin-adapters.test.mjs

@@ -42,6 +42,16 @@ const adapters = [
     expectedBaseUrl: 'https://api.deepseek.com',
     expectedApiKey: 'ds-key',
   },
+  {
+    name: 'nvidia-nim-api',
+    apiMode: {
+      groupName: 'nvidiaNimApiModelKeys',
+      itemName: 'nvidiaNim_nemotron_3_super',
+    },
+    providerId: 'nvidia-nim',
+    expectedBaseUrl: 'https://integrate.api.nvidia.com/v1',
+    expectedApiKey: 'nv-key',
+  },
   {
     name: 'moonshot-api',
     apiMode: { groupName: 'moonshotApiModelKeys', itemName: 'moonshot_kimi_latest' },