feat(providers): add CloudGPU

README.md

@@ -128,7 +128,7 @@ preview `ollama launch webbrain --model <model>` handoff. Details:
 **Cloud APIs** — OpenAI, Anthropic Claude, Google Gemini, Azure OpenAI, AWS
 Bedrock, Mistral, DeepSeek, xAI Grok, MiniMax, Kimi, Qwen, z.ai GLM, Groq,
 Together, Cloudflare, Nvidia NIM, Hugging Face, Fireworks, OpenRouter, and more.
-The Providers page ships **119 built-in provider cards on Chromium** (118 on Firefox),
+The Providers page ships **120 built-in provider cards on Chromium** (119 on Firefox),
 including an endpoint-free local WebGPU option with the tested LFM2.5 2.6B
 preset and an experimental custom Hugging Face ONNX repository option —
 see the [full catalog](docs/providers-and-models.md#extended-provider-catalog).

docs/providers-and-models.md

@@ -510,11 +510,11 @@ times the standard rate.
 
 ### Extended provider catalog
 
-WebBrain also ships 85 disabled-by-default provider cards. Most are sourced
+WebBrain also ships 86 disabled-by-default provider cards. Most are sourced
 from the OpenCode provider catalog snapshot at commit
 `62e4641235d7847dadc60da37cca8a023dd54fc1`; provider-specific additions use
 their official API documentation. Together with the original cards, the
-Providers page lists **119 built-in providers on Chromium** and **118 on Firefox**; the
+Providers page lists **120 built-in providers on Chromium** and **119 on Firefox**; the
 difference is the Chromium-only in-browser WebGPU runtime.
 
 | IDs                                                                                                                                                                                                                                                                                                                        |
@@ -523,6 +523,7 @@ difference is the Chromium-only in-browser WebGPU runtime.
 | `freebuff2api`, `google-vertex`, `google-vertex-anthropic`, `helicone`, `iflowcn`, `inception`, `inference`, `io-net`, `jiekou`, `kilo`, `kimi-for-coding`, `kuae-cloud-coding-plan`, `llama`, `lucidquery`, `meganova`, `minimax-cn-coding-plan`, `minimax-coding-plan`, `moark`, `modelscope`, `morph`                       |
 | `nano-gpt`, `nearai`, `nebius`, `nova`, `novita-ai`, `ollama-cloud`, `opencode`, `opencode-go`, `opper`, `orcarouter`, `ovhcloud`, `perplexity`, `perplexity-agent`, `poe`, `pollinations`, `privatemode-ai`, `qihang-ai`, `qiniu-ai`, `requesty`, `runcrate`, `scaleway`, `siliconflow`, `siliconflow-cn`, `stackit`        |
 | `stepfun`, `submodel`, `synthetic`, `tencent-coding-plan`, `upstage`, `v0`, `venice`, `vercel`, `vivgrid`, `vultr`, `wandb`, `xiaomi`, `zai-coding-plan`, `zenmux`, `zhipuai`, `zhipuai-coding-plan`                                                                                                                       |
+| `cloudgpu`                                                                                                                                                                                                                                                                                                                     |
 
 Most use the OpenAI-compatible Chat Completions contract and bearer API keys.
 The exceptions are:
@@ -646,6 +647,18 @@ through the existing OpenAI-compatible path. The card starts with vision off
 because the default model is text-only; enable it for a vision model such as
 `moonshotai/Kimi-K3`.
 
+#### CloudGPU
+
+[CloudGPU](https://cloudgpu.app) serves DeepSeek, Kimi, GLM, MiniMax, Qwen and
+more through an OpenAI-compatible API at `https://cloudgpu.app/v1`. The live
+model list is at `https://cloudgpu.app/v1/models`.
+
+In WebBrain, open **Providers -> LLM -> CloudGPU**. Paste an API key into the
+card, keep the default `deepseek-v4-flash` model (131072 tokens), then click
+**Test Connection**. Interactive Ask streaming and tool calls work through the
+existing OpenAI-compatible path. The card starts with vision off because the
+default model is text-only.
+
 ### Ask response streaming
 
 Providers with `supportsAskStreaming` stream visible text during interactive

src/chrome/icons/providers/SOURCES.md

@@ -21,6 +21,7 @@ Most brand marks come from [Lobe Icons](https://github.com/lobehub/lobe-icons)
 | nearai | WebBrain-authored simple N mark (not an official NEAR brand asset) |
 | cheaperinference | Cheaper Inference logo mark (cheaperinference.com) |
 | opper | Opper official logo mark (opper.ai) |
+| cloudgpu | CloudGPU logo mark (cloudgpu.app) |
 | runcrate | Runcrate logo mark (runcrate.ai) |
 
 Brand trademarks remain property of their owners; used only for provider identification in the settings UI.

src/firefox/icons/providers/SOURCES.md

@@ -21,6 +21,7 @@ Most brand marks come from [Lobe Icons](https://github.com/lobehub/lobe-icons)
 | nearai | WebBrain-authored simple N mark (not an official NEAR brand asset) |
 | cheaperinference | Cheaper Inference logo mark (cheaperinference.com) |
 | opper | Opper official logo mark (opper.ai) |
+| cloudgpu | CloudGPU logo mark (cloudgpu.app) |
 | runcrate | Runcrate logo mark (runcrate.ai) |
 
 Brand trademarks remain property of their owners; used only for provider identification in the settings UI.

test/provider-catalog-integrations.mjs

@@ -25,6 +25,17 @@ for (const browser of ['chrome', 'firefox']) {
   const { providerIconUrl } = await import(pathToFileURL(path.join(prefix, 'ui', 'provider-icons.js')).href);
   assert.ok(providerIconUrl('api-route').endsWith('/api-route.png'));
 
+  const cloudgpu = ADDITIONAL_PROVIDER_DEFAULTS.cloudgpu;
+  assert.equal(cloudgpu.type, 'openai', `${browser}: CloudGPU must use the OpenAI-compatible adapter`);
+  assert.equal(cloudgpu.baseUrl, 'https://cloudgpu.app/v1', `${browser}: CloudGPU endpoint changed`);
+  assert.equal(cloudgpu.model, 'deepseek-v4-flash');
+  assert.equal(cloudgpu.contextWindow, 131072);
+  assert.equal(cloudgpu.enabled, false);
+  assert.equal(cloudgpu.apiKey, '');
+  assert.equal(cloudgpu.supportsTools, true);
+  assert.equal(cloudgpu.supportsAskStreaming, true);
+  assert.ok(providerIconUrl('cloudgpu').endsWith('/cloudgpu.svg'));
+
   const runcrate = ADDITIONAL_PROVIDER_DEFAULTS.runcrate;
   assert.equal(runcrate.type, 'openai', `${browser}: Runcrate must use the OpenAI-compatible adapter`);
   assert.equal(runcrate.baseUrl, 'https://api.runcrate.ai/v1', `${browser}: Runcrate endpoint changed`);

test/run.js

@@ -76986,6 +76986,7 @@ test('extended provider catalog is complete, mirrored, safe, and excluded-provid
   const expectedIds = `
     api-route 302ai abacus aihubmix alibaba-coding-plan alibaba-coding-plan-cn
     atlascloud azure-cognitive-services bailing baseten berget cerebras cheaperinference chutes clarifai
+    cloudgpu
     cloudferro-sherlock cohere cortecs deepinfra digitalocean dinference drun
     evroc fastrouter freebuff2api friendli google-vertex google-vertex-anthropic helicone
     iflowcn inception inference io-net jiekou kilo kimi-for-coding
@@ -76999,7 +77000,7 @@ test('extended provider catalog is complete, mirrored, safe, and excluded-provid
   `.trim().split(/\s+/);
   const excluded = ['github-models', 'github-copilot', 'gitlab', 'sap-ai-core'];
 
-  assert.equal(expectedIds.length, 85);
+  assert.equal(expectedIds.length, 86);
   assert.deepEqual(ProviderCatalogCh.ADDITIONAL_PROVIDER_IDS, expectedIds);
   assert.deepEqual(ProviderCatalogFx.ADDITIONAL_PROVIDER_IDS, expectedIds);
   assert.deepEqual(
@@ -77014,8 +77015,8 @@ test('extended provider catalog is complete, mirrored, safe, and excluded-provid
   ]) {
     const defaults = new PM()._defaultConfigs();
     // Includes the sidepanel-only WebBrain Compass XL runtime profile, which is
-    // deliberately omitted from the 119/118 Settings provider cards.
-    const expectedDefaultCount = label === 'chrome' ? 120 : 119;
+    // deliberately omitted from the 120/119 Settings provider cards.
+    const expectedDefaultCount = label === 'chrome' ? 121 : 120;
     assert.equal(
       Object.keys(defaults).length,
       expectedDefaultCount,
@@ -77215,6 +77216,30 @@ test('extended provider catalog is complete, mirrored, safe, and excluded-provid
     ProviderCatalogCh.ADDITIONAL_PROVIDER_UI['api-route'].suggestions,
     ['gpt-6.1-sol'],
   );
+  assert.deepEqual(
+    {
+      baseUrl: ProviderCatalogCh.ADDITIONAL_PROVIDER_DEFAULTS.cloudgpu.baseUrl,
+      model: ProviderCatalogCh.ADDITIONAL_PROVIDER_DEFAULTS.cloudgpu.model,
+      contextWindow: ProviderCatalogCh.ADDITIONAL_PROVIDER_DEFAULTS.cloudgpu.contextWindow,
+      supportsVision: ProviderCatalogCh.ADDITIONAL_PROVIDER_DEFAULTS.cloudgpu.supportsVision,
+      supportsTools: ProviderCatalogCh.ADDITIONAL_PROVIDER_DEFAULTS.cloudgpu.supportsTools,
+      supportsAskStreaming: ProviderCatalogCh.ADDITIONAL_PROVIDER_DEFAULTS.cloudgpu.supportsAskStreaming,
+      apiKeyUrl: ProviderCatalogCh.ADDITIONAL_PROVIDER_DEFAULTS.cloudgpu.apiKeyUrl,
+    },
+    {
+      baseUrl: 'https://cloudgpu.app/v1',
+      model: 'deepseek-v4-flash',
+      contextWindow: 131072,
+      supportsVision: false,
+      supportsTools: true,
+      supportsAskStreaming: true,
+      apiKeyUrl: 'https://cloudgpu.app',
+    },
+  );
+  assert.deepEqual(
+    ProviderCatalogCh.ADDITIONAL_PROVIDER_UI.cloudgpu.suggestions,
+    ['deepseek-v4-flash', 'deepseek-v4-pro', 'glm-5.3', 'kimi-k3', 'qwen3.5-397b'],
+  );
   assert.deepEqual(
     {
       baseUrl: ProviderCatalogCh.ADDITIONAL_PROVIDER_DEFAULTS.runcrate.baseUrl,