feat(provider): add AI/ML API provider

.env.example

@@ -30,6 +30,12 @@
 #     OPENAI_MODEL=gpt-4o
 #     OPENAI_BASE_URL=https://api.openai.com/v1      (optional)
 #
+#   Option 2a — AI/ML API:
+#     CLAUDE_CODE_USE_OPENAI=1
+#     AIMLAPI_API_KEY=your-aimlapi-key-here
+#     OPENAI_BASE_URL=https://api.aimlapi.com/v1
+#     OPENAI_MODEL=gpt-4o
+#
 #   Option 3 — Google Gemini:
 #     CLAUDE_CODE_USE_GEMINI=1
 #     GEMINI_API_KEY=your-gemini-key-here
@@ -150,6 +156,19 @@ ANTHROPIC_API_KEY=sk-ant-your-key-here
 # OPENAI_BASE_URL=https://api.openai.com/v1
 
 
+# -----------------------------------------------------------------------------
+# Option 2a: AI/ML API
+# -----------------------------------------------------------------------------
+# AI/ML API provides an OpenAI-compatible endpoint for 400+ models across
+# chat, code, image, video, voice, embeddings, and more. OpenClaude uses
+# chat/tool-capable models through the chat completions API.
+#
+# CLAUDE_CODE_USE_OPENAI=1
+# AIMLAPI_API_KEY=your-aimlapi-key-here
+# OPENAI_BASE_URL=https://api.aimlapi.com/v1
+# OPENAI_MODEL=gpt-4o
+
+
 # -----------------------------------------------------------------------------
 # Option 3: Google Gemini
 # -----------------------------------------------------------------------------

README.md

@@ -70,6 +70,34 @@ $env:OPENAI_MODEL="gpt-4o"
 openclaude
 ```
 
+### Fastest AI/ML API setup
+
+AI/ML API provides 400+ models behind an OpenAI-compatible API. OpenClaude uses chat/tool-capable models through `https://api.aimlapi.com/v1/chat/completions`.
+
+macOS / Linux:
+
+```bash
+export CLAUDE_CODE_USE_OPENAI=1
+export AIMLAPI_API_KEY=your-aimlapi-key-here
+export OPENAI_BASE_URL=https://api.aimlapi.com/v1
+export OPENAI_MODEL=gpt-4o
+
+openclaude
+```
+
+Windows PowerShell:
+
+```powershell
+$env:CLAUDE_CODE_USE_OPENAI="1"
+$env:AIMLAPI_API_KEY="your-aimlapi-key-here"
+$env:OPENAI_BASE_URL="https://api.aimlapi.com/v1"
+$env:OPENAI_MODEL="gpt-4o"
+
+openclaude
+```
+
+You can also run `/provider`, choose `AI/ML API`, and paste your AI/ML API key. Browse available models at [aimlapi.com/models](https://aimlapi.com/models) and provider docs at [docs.aimlapi.com](https://docs.aimlapi.com/).
+
 ### Fastest local Ollama setup
 
 macOS / Linux:
@@ -112,13 +140,15 @@ Beginner-friendly guides:
 
 Advanced and source-build guides:
 
+- [AI/ML API Setup](docs/aimlapi-setup.md)
 - [Advanced Setup](docs/advanced-setup.md)
 - [Android Install](ANDROID_INSTALL.md)
 
 ## Supported Providers
 
 | Provider | Setup Path | Notes |
 | --- | --- | --- |
+| AI/ML API | `/provider` or env vars | First-class OpenAI-compatible setup for AI/ML API's 400+ model catalog; OpenClaude uses chat/tool-capable models |
 | OpenAI-compatible | `/provider` or env vars | Works with OpenAI, OpenRouter, DeepSeek, Groq, Mistral, LM Studio, and other compatible `/v1` servers |
 | Gemini | `/provider` or env vars | Supports API key, access token, or local ADC workflow on current `main` |
 | GitHub Models | `/onboard-github` | Interactive onboarding with saved credentials |

docs/aimlapi-setup.md

@@ -0,0 +1,35 @@
+# AI/ML API Setup
+
+OpenClaude can run through AI/ML API as a first-class OpenAI-compatible provider. AI/ML API exposes `https://api.aimlapi.com/v1` and OpenClaude sends chat requests to `https://api.aimlapi.com/v1/chat/completions`.
+
+## Setup with `/provider`
+
+1. Start OpenClaude.
+2. Run `/provider`.
+3. Choose `AI/ML API`.
+4. Paste your AI/ML API key.
+5. Keep the default model `gpt-4o` or enter another chat-capable model ID.
+
+## Setup with environment variables
+
+macOS / Linux:
+
+```bash
+export CLAUDE_CODE_USE_OPENAI=1
+export AIMLAPI_API_KEY=your-aimlapi-key-here
+export OPENAI_BASE_URL=https://api.aimlapi.com/v1
+export OPENAI_MODEL=gpt-4o
+```
+
+Windows PowerShell:
+
+```powershell
+$env:CLAUDE_CODE_USE_OPENAI="1"
+$env:AIMLAPI_API_KEY="your-aimlapi-key-here"
+$env:OPENAI_BASE_URL="https://api.aimlapi.com/v1"
+$env:OPENAI_MODEL="gpt-4o"
+```
+
+OpenClaude uses chat/tool-capable models for coding-agent workflows. AI/ML API also offers other modalities such as image, video, voice, music, embeddings, OCR, and 3D generation; those models are available through AI/ML API but are outside OpenClaude's core chat/tool loop.
+
+Browse models at [aimlapi.com/models](https://aimlapi.com/models) and read provider docs at [docs.aimlapi.com](https://docs.aimlapi.com/).

src/commands/provider/provider.tsx

@@ -23,6 +23,10 @@ import {
   resolveCodexApiCredentials,
   resolveProviderRequest,
 } from '../../services/api/providerConfig.js'
+import {
+  AIMLAPI_LABEL,
+  isAimlapiBaseUrl,
+} from '../../providers/aimlapi/index.js'
 import {
   applySavedProfileToCurrentSession as applySharedProfileToCurrentSession,
   buildCodexOAuthProfileEnv as buildSharedCodexOAuthProfileEnv,
@@ -207,10 +211,10 @@ export function getProviderWizardDefaults(
     sanitizeProviderConfigValue(processEnv.GEMINI_MODEL, secretSource) ||
     DEFAULT_GEMINI_MODEL
   const safeMistralModel =
-    sanitizeProviderConfigValue(processEnv.MISTRAL_MODEL, processEnv) ||
+    sanitizeProviderConfigValue(processEnv.MISTRAL_MODEL, secretSource) ||
     DEFAULT_MISTRAL_MODEL
   const safeMistralBaseUrl =
-    sanitizeProviderConfigValue(processEnv.MISTRAL_BASE_URL, processEnv) ||
+    sanitizeProviderConfigValue(processEnv.MISTRAL_BASE_URL, secretSource) ||
     DEFAULT_MISTRAL_BASE_URL
 
   return {
@@ -287,6 +291,8 @@ export function buildCurrentProviderSummary(options?: {
     let providerLabel = 'OpenAI-compatible'
     if (request.transport === 'codex_responses') {
       providerLabel = 'Codex'
+    } else if (isAimlapiBaseUrl(request.baseUrl)) {
+      providerLabel = AIMLAPI_LABEL
     } else if (isLocalProviderUrl(request.baseUrl)) {
       providerLabel = getLocalOpenAICompatibleProviderLabel(request.baseUrl)
     }

src/components/ProviderManager.test.tsx

@@ -105,6 +105,7 @@ async function waitForCondition(
 // Order matches ProviderManager.renderPresetSelection() when
 // canUseCodexOAuth === true (default in mocked tests).
 const PRESET_ORDER = [
+  'AI/ML API',
   'Alibaba Coding Plan',
   'Alibaba Coding Plan (China)',
   'Anthropic',
@@ -437,6 +438,23 @@ test('ProviderManager resolves GitHub virtual provider from async storage withou
   expect(asyncRead).toHaveBeenCalled()
 })
 
+test('ProviderManager preset picker includes AI/ML API', async () => {
+  mockProviderManagerDependencies(
+    () => undefined,
+    async () => undefined,
+  )
+
+  const nonce = `${Date.now()}-${Math.random()}`
+  const { ProviderManager } = await import(`./ProviderManager.js?ts=${nonce}`)
+  const output = await renderProviderManagerFrame(ProviderManager, {
+    mode: 'first-run',
+    waitForOutput: frame =>
+      frame.includes('Set up provider') && frame.includes('AI/ML API'),
+  })
+
+  expect(output).toContain('AI/ML API')
+})
+
 test('ProviderManager avoids first-frame false negative while stored-token lookup is pending', async () => {
   delete process.env.CLAUDE_CODE_USE_GITHUB
   delete process.env.GITHUB_TOKEN

src/components/ProviderManager.tsx

@@ -42,6 +42,9 @@ import {
   type AtomicChatReadiness,
   type OllamaGenerationReadiness,
 } from '../utils/providerDiscovery.js'
+import {
+  AIMLAPI_PROVIDER_PRESET_OPTION,
+} from '../providers/aimlapi/index.js'
 import {
   rankOllamaModels,
   recommendOllamaModel,
@@ -1254,6 +1257,7 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
     // should always find known providers first. `Skip for now` (first-run
     // only) comes last, after Custom.
     const options = [
+      AIMLAPI_PROVIDER_PRESET_OPTION,
       {
         value: 'dashscope-intl',
         label: 'Alibaba Coding Plan',

src/providers/aimlapi/index.test.ts

@@ -0,0 +1,141 @@
+import { describe, expect, test } from 'bun:test'
+import {
+  AIMLAPI_ATTRIBUTION_HEADERS,
+  AIMLAPI_DEFAULT_BASE_URL,
+  AIMLAPI_DEFAULT_MODEL,
+  AIMLAPI_LABEL,
+  AIMLAPI_PROVIDER_ID,
+  AIMLAPI_PROVIDER_PRESET_OPTION,
+  getAimlapiApiKey,
+  getAimlapiAttributionHeaders,
+  getAimlapiOpenAICompatibleApiKey,
+  getAimlapiPresetDefaults,
+  hasAimlapiApiKey,
+  isAimlapiBaseUrl,
+  mapAimlapiModelCatalog,
+  syncAimlapiOpenAIEnv,
+} from './index.js'
+
+describe('AI/ML API provider module', () => {
+  test('detects only api.aimlapi.com base URLs', () => {
+    expect(isAimlapiBaseUrl('https://api.aimlapi.com/v1')).toBe(true)
+    expect(isAimlapiBaseUrl('https://api.aimlapi.com/v1/')).toBe(true)
+    expect(isAimlapiBaseUrl('https://example.com/api.aimlapi.com/v1')).toBe(false)
+    expect(isAimlapiBaseUrl('not a url')).toBe(false)
+    expect(isAimlapiBaseUrl(undefined)).toBe(false)
+  })
+
+  test('exposes preset defaults and picker option metadata', () => {
+    expect(getAimlapiPresetDefaults({
+      AIMLAPI_API_KEY: 'aiml-key',
+      OPENAI_API_KEY: 'openai-key',
+    })).toEqual({
+      provider: 'openai',
+      name: AIMLAPI_LABEL,
+      baseUrl: AIMLAPI_DEFAULT_BASE_URL,
+      model: AIMLAPI_DEFAULT_MODEL,
+      apiKey: 'aiml-key',
+      requiresApiKey: true,
+    })
+
+    expect(AIMLAPI_PROVIDER_PRESET_OPTION).toEqual({
+      value: AIMLAPI_PROVIDER_ID,
+      label: AIMLAPI_LABEL,
+      description: 'AI/ML API OpenAI-compatible endpoint',
+    })
+  })
+
+  test('resolves provider-specific auth without affecting other base URLs', () => {
+    const env = {
+      AIMLAPI_API_KEY: 'aiml-key',
+      OPENAI_API_KEY: 'openai-key',
+    }
+
+    expect(getAimlapiApiKey(env)).toBe('aiml-key')
+    expect(getAimlapiApiKey({ OPENAI_API_KEY: 'openai-key' })).toBe('openai-key')
+    expect(getAimlapiOpenAICompatibleApiKey(
+      'https://api.aimlapi.com/v1',
+      env,
+    )).toBe('aiml-key')
+    expect(getAimlapiOpenAICompatibleApiKey(
+      'https://openrouter.ai/api/v1',
+      env,
+    )).toBeUndefined()
+    expect(hasAimlapiApiKey('https://api.aimlapi.com/v1', env)).toBe(true)
+  })
+
+  test('adds attribution headers only for AI/ML API', () => {
+    expect(getAimlapiAttributionHeaders('https://api.aimlapi.com/v1')).toEqual(
+      AIMLAPI_ATTRIBUTION_HEADERS,
+    )
+    expect(getAimlapiAttributionHeaders('https://api.openai.com/v1')).toEqual({})
+  })
+
+  test('syncs AIMLAPI_API_KEY into OPENAI_API_KEY only for AI/ML API env', () => {
+    const aimlEnv = {
+      OPENAI_BASE_URL: 'https://api.aimlapi.com/v1',
+      AIMLAPI_API_KEY: 'aiml-key',
+    }
+    syncAimlapiOpenAIEnv(aimlEnv)
+    expect(aimlEnv.OPENAI_API_KEY).toBe('aiml-key')
+
+    const existingOpenAIEnv = {
+      OPENAI_BASE_URL: 'https://api.aimlapi.com/v1',
+      AIMLAPI_API_KEY: 'aiml-key',
+      OPENAI_API_KEY: 'openai-key',
+    }
+    syncAimlapiOpenAIEnv(existingOpenAIEnv)
+    expect(existingOpenAIEnv.OPENAI_API_KEY).toBe('openai-key')
+
+    const otherEnv = {
+      OPENAI_BASE_URL: 'https://api.openai.com/v1',
+      AIMLAPI_API_KEY: 'aiml-key',
+    }
+    syncAimlapiOpenAIEnv(otherEnv)
+    expect(otherEnv.OPENAI_API_KEY).toBeUndefined()
+  })
+
+  test('maps chat-completions models with metadata and deduplication', () => {
+    expect(mapAimlapiModelCatalog({
+      data: [
+        {
+          id: 'gpt-4o',
+          type: 'openai/chat-completions',
+          info: {
+            name: 'GPT 4o',
+            developer: 'OpenAI',
+            contextLength: 128000,
+          },
+        },
+        {
+          id: 'gpt-4o',
+          type: 'openai/chat-completions',
+          info: { name: 'Duplicate GPT 4o' },
+        },
+        {
+          id: 'image-model',
+          type: 'openai/images',
+        },
+        {
+          id: 'deepseek-chat',
+          type: 'openai/chat-completions',
+          info: {
+            name: 'DeepSeek Chat',
+            developer: 'DeepSeek',
+          },
+        },
+      ],
+    })).toEqual([
+      {
+        value: 'gpt-4o',
+        label: 'GPT 4o',
+        description: 'OpenAI - 128000 context',
+      },
+      {
+        value: 'deepseek-chat',
+        label: 'DeepSeek Chat',
+        description: 'DeepSeek',
+      },
+    ])
+  })
+})