Skip to content

Open-source AI browser agent for Chrome and Firefox (monorepo) 🧠

Add Osaurus provider support

Emre Sokullu142dc75dc137
diff --git a/README.md b/README.md
index cde1ab715..af846998b 100644
--- a/README.md
+++ b/README.md
@@ -89,13 +89,15 @@ vllm serve your-model --port 8000 # vLLM β†’ :8000/v1
python -m sglang.launch_server --model-path your-model --port 30000
```
-LM Studio (`:1234/v1`), Jan (`:1337/v1`), LocalAI (`:8080/v1`), and GPT4All
+LM Studio (`:1234/v1`), Osaurus (`:1337/v1`), Jan (`:1337/v1`), LocalAI (`:8080/v1`), and GPT4All
(`:4891/v1`) work the same way. A generic **Local OpenAI-compatible Proxy** card
also supports authenticated loopback gateways such as CLIProxyAPI; see the
[secure subscription proxy setup](docs/providers-and-models.md#subscription-proxy-example-cliproxyapi).
**Unsloth Studio (Local)** uses the same OpenAI-compatible path with the Studio
port and `sk-unsloth-` API key configured by the user; see the
[Unsloth Studio setup](docs/providers-and-models.md#unsloth-studio).
+**Osaurus (Local)** connects to the Mac's Osaurus server with model discovery,
+streaming, and tool calls; see the [Osaurus setup](docs/providers-and-models.md#osaurus).
Load a model with **at least a 16k-token context window** β€” 8k works only
with the Compact tier, and 4k is too small for the system prompt plus tool
schemas. WebBrain auto-detects the real window for llama.cpp, Ollama, and LM
diff --git a/docs/providers-and-models.md b/docs/providers-and-models.md
index d5b6d3db0..a40656887 100644
--- a/docs/providers-and-models.md
+++ b/docs/providers-and-models.md
@@ -87,6 +87,7 @@ class BaseLLMProvider {
| `llamacpp` | `llamacpp` | local | (loaded model) | Auto metadata / override |
| `ollama` | `openai` | local | (loaded model) | Auto via `/api/show` / override |
| `lmstudio` | `openai` | local | (loaded model) | Auto metadata / override |
+| `osaurus` | `openai` | local | (required) | Off / manual toggle |
| `jan` | `openai` | local | (loaded model) | Yes (default on) |
| `vllm` | `openai` | local | (loaded model) | Yes (default on) |
| `sglang` | `openai` | local | (loaded model) | Yes (default on) |
@@ -121,7 +122,7 @@ WebBrain also ships 79 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, Settings
-contains **110 built-in providers on Chromium** and **109 on Firefox**; the
+contains **111 built-in providers on Chromium** and **110 on Firefox**; the
difference is the Chromium-only in-browser WebGPU runtime.
| IDs |
@@ -250,13 +251,14 @@ URL, localhost server, or OpenAI-compatible endpoint. Firefox does not expose
the card because its build does not package the Chromium MV3 offscreen/WebGPU
runtime.
-Ten local endpoint providers are enabled by default. The model runtimes need no
+Eleven local endpoint providers are enabled by default. The model runtimes need no
API key unless the server was started with auth; Unsloth Studio and the generic
proxy card require their configured client keys:
- **llama.cpp**: `http://localhost:8080` β€” runs `llama-server -m model.gguf`
- **Ollama**: `http://localhost:11434/v1` β€” `ollama serve`, or `ollama launch webbrain --model <model>`
- **LM Studio**: `http://localhost:1234/v1` β€” LM Studio's local inference server
+- **Osaurus**: `http://127.0.0.1:1337/v1` β€” `osaurus serve --port 1337` on macOS
- **Jan**: `http://localhost:1337/v1` β€” Jan's local OpenAI-compatible API server
- **vLLM**: `http://localhost:8000/v1` β€” vLLM's OpenAI-compatible server
- **SGLang**: `http://localhost:30000/v1` β€” SGLang's OpenAI-compatible server
@@ -267,6 +269,24 @@ proxy card require their configured client keys:
- **Unsloth Studio**: `http://127.0.0.1:8888/v1` by default, with a configurable
port β€” Studio's API URL, loaded model, and generated API key are required
+#### Osaurus
+
+Start the Osaurus server on your Mac, then open **Settings β†’ Providers β†’
+Osaurus (Local)**. Keep `http://127.0.0.1:1337/v1` as the server URL, click
+**Load Models**, select a model, and click **Test Connection**. A model ID is
+required; discovery uses Osaurus's OpenAI-compatible `/v1/models` endpoint.
+Chat, tool calls, and interactive Ask streaming use `/v1/chat/completions`.
+See the [official API guide](https://github.com/osaurus-ai/osaurus/blob/main/docs/OpenAI_API_GUIDE.md).
+
+Local-only loopback access needs no API key. If Osaurus requires authentication
+(including network exposure or relay access), enter its access key under the
+card's **Advanced β†’ API key** section. Osaurus and Jan both default to port
+1337; use different ports if running both and update the corresponding server
+URL. Set **Context window** to the selected model's actual runtime limit.
+Vision starts off; enable **Supports vision** only for a model that accepts
+images. Osaurus can also route to cloud providers, in which case that upstream
+provider receives the request.
+
#### Unsloth Studio
Install or open [Unsloth Studio](https://unsloth.ai/docs/get-started/install),
@@ -336,7 +356,7 @@ OLLAMA_ORIGINS="chrome-extension://*,moz-extension://*" ./ollama serve
**Streaming.** Local streaming is primarily a runtime/server capability, not a
property of the GGUF or other model weights. Interactive Ask streaming is
-enabled for llama.cpp, Ollama, LM Studio, Jan, vLLM, SGLang, current LocalAI,
+enabled for llama.cpp, Ollama, LM Studio, Osaurus, Jan, vLLM, SGLang, current LocalAI,
and Unsloth Studio
through their OpenAI-compatible Chat Completions endpoints. Each parser requires
`[DONE]`; safe network/read, malformed-frame, and premature-EOF failures
diff --git a/src/chrome/icons/providers/SOURCES.md b/src/chrome/icons/providers/SOURCES.md
index c54fdfa65..157f46436 100644
--- a/src/chrome/icons/providers/SOURCES.md
+++ b/src/chrome/icons/providers/SOURCES.md
@@ -8,6 +8,7 @@ Most brand marks come from [Lobe Icons](https://github.com/lobehub/lobe-icons)
| openai, anthropic, gemini, ollama, lmstudio, openrouter, groq, mistral, deepseek, xai, huggingface, fireworks, together, cloudflare, nvidia, azure_openai, aws_bedrock, minimax, kimi, alibaba (qwen), vllm, pollinations | Lobe Icons |
| llamacpp | Jan app asset (brand mark used for llama.cpp) |
| jan | Jan official app icon |
+| osaurus | [Osaurus official logo](https://github.com/osaurus-ai/osaurus/blob/main/App/osaurus/Assets.xcassets/osaurus-logo.imageset/osaurus-logo-black.svg) (`osaurus-ai/osaurus`, MIT) |
| localai | LocalAI official logo (`mudler/LocalAI`) |
| gpt4all | Simple mark (not an official GPT4All brand asset) |
| local_openai_proxy, webgpu | WebBrain-authored generic local-runtime marks |
diff --git a/src/chrome/icons/providers/osaurus.svg b/src/chrome/icons/providers/osaurus.svg
new file mode 100644
index 000000000..698845466
--- /dev/null
+++ b/src/chrome/icons/providers/osaurus.svg
@@ -0,0 +1,24 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 200 200" fill="none" preserveAspectRatio="xMidYMid meet">
+ <defs>
+ <style>
+ .st0 {
+ fill: none;
+ stroke: #000000;
+ stroke-miterlimit: 10;
+ stroke-width: 16px;
+ }
+ </style>
+ </defs>
+ <g>
+ <g>
+ <line class="st0" x1="98.2" y1="191.2" x2="98.2" y2="166" />
+ <path class="st0" d="M61.6,164.1s-21.5,10.4-49.2-17.1v-.2c39.1-.2,60.1-35.9,60.1-35.9" />
+ </g>
+ <path
+ class="st0"
+ d="M61.6,191.2v-35.8c0-42.8,16.7-63.6,58.1-63.6v-43.9s-.5-29.4,34.1-29.7,34.4,15,34.4,33.7c0,18.8-9.4,27.4-22.5,32.1v29.8c0,14.2-6.2,27.8-16.9,37.2l-17.1,14.9v25.2"
+ />
+ </g>
+ <circle cx="155.8" cy="51.4" r="7.8" fill="#000000" stroke="none" />
+</svg>
diff --git a/src/chrome/src/background.js b/src/chrome/src/background.js
index 9237c4d4d..6ea8d2c59 100644
--- a/src/chrome/src/background.js
+++ b/src/chrome/src/background.js
@@ -4119,7 +4119,9 @@ async function handleMessage(msg, sender) {
}
case 'list_provider_models': {
- return await providerManager.listProviderModels(msg.providerId);
+ return await providerManager.listProviderModels(msg.providerId, {
+ detectServerIdentity: msg.detectServerIdentity === true,
+ });
}
case 'list_ollama_models': {
diff --git a/src/chrome/src/providers/manager.js b/src/chrome/src/providers/manager.js
index 7127b11c2..421b833b8 100644
--- a/src/chrome/src/providers/manager.js
+++ b/src/chrome/src/providers/manager.js
@@ -79,7 +79,7 @@ import {
const WEBBRAIN_CLOUD_PROVIDER_ID = 'webbrain_cloud';
const WEBBRAIN_CLOUD_PROVIDER_LABEL = 'WebBrain Compass';
const DUPLICATE_PROVIDER_SUFFIX = '__duplicate';
-const LOCAL_MODEL_LIST_PROVIDER_IDS = ['llamacpp', 'ollama', 'lmstudio', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth'];
+const LOCAL_MODEL_LIST_PROVIDER_IDS = ['llamacpp', 'ollama', 'lmstudio', 'osaurus', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth'];
const WEBBRAIN_CLOUD_CONTEXT_WINDOW = 1000000;
const WEBBRAIN_CLOUD_LEGACY_CONTEXT_WINDOW = 256000;
const WEBBRAIN_DEVICE_GUID_KEY = 'webbrainDeviceGuid';
@@ -552,6 +552,20 @@ export class ProviderManager {
visionDetection: null,
enabled: true,
},
+ osaurus: {
+ type: 'openai',
+ category: 'local',
+ label: 'Osaurus (Local)',
+ providerName: 'osaurus',
+ requiresModel: true,
+ baseUrl: 'http://127.0.0.1:1337/v1',
+ model: '',
+ contextWindow: 16384,
+ apiKey: '',
+ supportsAskStreaming: true,
+ supportsVision: false,
+ enabled: true,
+ },
jan: {
type: 'openai',
category: 'local',
@@ -1262,7 +1276,7 @@ export class ProviderManager {
static categoryFor(id, config) {
if (config && config.category) return config.category;
if (config?.type === 'llamacpp' || config?.type === 'webgpu') return 'local';
- if (['llamacpp', 'ollama', 'lmstudio', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth', 'webgpu'].includes(id)) return 'local';
+ if (['llamacpp', 'ollama', 'lmstudio', 'osaurus', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth', 'webgpu'].includes(id)) return 'local';
if (ROUTER_PROVIDER_IDS.includes(id)) return 'router';
return 'cloud';
}
@@ -2302,11 +2316,11 @@ export class ProviderManager {
/**
* Fetch selectable models for local providers. Ollama uses its native
- * /api/tags endpoint; llama.cpp, LM Studio, Jan, vLLM, SGLang, LocalAI,
+ * /api/tags endpoint; llama.cpp, LM Studio, Osaurus, Jan, vLLM, SGLang, LocalAI,
* GPT4All, Unsloth Studio, and generic local proxies use
* OpenAI-compatible /v1/models.
*/
- async listProviderModels(id) {
+ async listProviderModels(id, { detectServerIdentity = false } = {}) {
const provider = this.providers.get(id);
if (!provider) return { ok: false, error: 'Provider not found' };
const definitionId = this._providerDefinitionId(id, provider.config);
@@ -2378,6 +2392,13 @@ export class ProviderManager {
continue;
}
const data = await res.json();
+ // Jan and Osaurus share port 1337. Successful OpenAI model discovery
+ // alone cannot identify which server answered an unconfigured card.
+ if (detectServerIdentity && provider.config.configured !== true
+ && ['jan', 'osaurus'].includes(definitionId)) {
+ const isOsaurus = await this._isOsaurusDiscoveryServer(data, candidate.requestBaseUrl, headers);
+ if (isOsaurus !== (definitionId === 'osaurus')) return { ok: true, models: [] };
+ }
const models = this._extractModelIds(definitionId, data);
const result = { ok: true, models };
if (candidate.configBaseUrl !== observedBaseUrl) {
@@ -2405,6 +2426,28 @@ export class ProviderManager {
return firstFailure || { ok: false, error: 'Failed to load models' };
}
+ async _isOsaurusDiscoveryServer(data, requestBaseUrl, headers) {
+ if (Array.isArray(data?.data) && data.data.some(model =>
+ String(model?.owned_by || '').trim().toLowerCase() === 'osaurus'
+ )) return true;
+
+ // A cloud-only Osaurus catalog may contain only upstream owners. Its
+ // public root greeting identifies the server without depending on models.
+ const root = requestBaseUrl.replace(/\/v1$/i, '');
+ const controller = new AbortController();
+ const timer = setTimeout(() => controller.abort(), 1500);
+ try {
+ const response = await fetchWithFallback(`${root}/`, {
+ method: 'GET', headers, signal: controller.signal, timeoutMs: 1500,
+ });
+ return response.ok && (await response.text()).trim().startsWith('Osaurus Server is running!');
+ } catch {
+ return false;
+ } finally {
+ clearTimeout(timer);
+ }
+ }
+
async listOllamaModels(id) {
return this.listProviderModels(id);
}
diff --git a/src/chrome/src/ui/provider-icons.js b/src/chrome/src/ui/provider-icons.js
index 5356a8496..ebabe0823 100644
--- a/src/chrome/src/ui/provider-icons.js
+++ b/src/chrome/src/ui/provider-icons.js
@@ -15,6 +15,7 @@ export const PROVIDER_ICON_FILES = {
llamacpp: 'llamacpp.svg',
ollama: 'ollama.svg',
lmstudio: 'lmstudio.svg',
+ osaurus: 'osaurus.svg',
jan: 'jan.png',
vllm: 'vllm.svg',
sglang: 'sglang.png',
@@ -83,6 +84,7 @@ export const PROVIDER_SHORT_LABELS = {
llamacpp: 'llama.cpp',
ollama: 'Ollama',
lmstudio: 'LM Studio',
+ osaurus: 'Osaurus',
jan: 'Jan',
vllm: 'vLLM',
sglang: 'SGLang',
diff --git a/src/chrome/src/ui/settings.js b/src/chrome/src/ui/settings.js
index 3769bd7e1..304a34928 100644
--- a/src/chrome/src/ui/settings.js
+++ b/src/chrome/src/ui/settings.js
@@ -3129,6 +3129,16 @@ function renderProviders() {
PROMPT_TIER_FIELD,
],
},
+ osaurus: {
+ fields: [
+ { key: 'baseUrl', labelKey: 'st.provider.field.server_url', type: 'text', placeholder: 'http://127.0.0.1:1337/v1' },
+ OPTIONAL_LOCAL_API_KEY_FIELD,
+ { key: 'model', labelKey: 'st.provider.field.model', type: 'text', placeholder: 'llama-3.2-3b-instruct' },
+ CONTEXT_WINDOW_FIELD,
+ { key: 'supportsVision', labelKey: 'st.provider.field.supports_vision', type: 'checkbox' },
+ PROMPT_TIER_FIELD,
+ ],
+ },
jan: {
fields: [
{ key: 'baseUrl', labelKey: 'st.provider.field.server_url', type: 'text', placeholder: 'http://localhost:1337/v1' },
@@ -3577,7 +3587,7 @@ function renderProviders() {
</div>
`;
} else {
- const localModelProviders = ['llamacpp', 'ollama', 'lmstudio', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth'];
+ const localModelProviders = ['llamacpp', 'ollama', 'lmstudio', 'osaurus', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth'];
const canLoadModels = localModelProviders.includes(definitionId) && field.key === 'model';
const listAttr = canLoadModels ? `list="models-${id}"` : '';
const datalistHTML = canLoadModels ? `<datalist id="models-${id}"></datalist>` : '';
diff --git a/src/chrome/src/ui/sidepanel.js b/src/chrome/src/ui/sidepanel.js
index 8ce807ac5..95df8476f 100644
--- a/src/chrome/src/ui/sidepanel.js
+++ b/src/chrome/src/ui/sidepanel.js
@@ -250,7 +250,7 @@ const pinCoachmarkDismissed = (async function initPinCoachmark() {
const localModels = document.getElementById('ob-local-models');
const localModelList = document.getElementById('ob-local-model-list');
const totalSteps = steps.length;
- const LOCAL_PROVIDER_ORDER = ['unsloth', 'local_openai_proxy', 'jan', 'lmstudio', 'ollama', 'llamacpp', 'vllm', 'sglang', 'localai', 'gpt4all'];
+ const LOCAL_PROVIDER_ORDER = ['unsloth', 'local_openai_proxy', 'jan', 'osaurus', 'lmstudio', 'ollama', 'llamacpp', 'vllm', 'sglang', 'localai', 'gpt4all'];
let current = 0;
let localScanStarted = false;
let localModelChoices = [];
@@ -462,7 +462,7 @@ const pinCoachmarkDismissed = (async function initPinCoachmark() {
// service-worker dynamic import in providers/manager.js, since fixed;
// the old 2.5s was also too tight for a cold proxy.)
const res = await withTimeout(
- sendToBackground('list_provider_models', { providerId }),
+ sendToBackground('list_provider_models', { providerId, detectServerIdentity: true }),
5000
);
if (res?.ok && Array.isArray(res.models)) {
diff --git a/src/firefox/icons/providers/SOURCES.md b/src/firefox/icons/providers/SOURCES.md
index 5af534379..d422323b6 100644
--- a/src/firefox/icons/providers/SOURCES.md
+++ b/src/firefox/icons/providers/SOURCES.md
@@ -8,6 +8,7 @@ Most brand marks come from [Lobe Icons](https://github.com/lobehub/lobe-icons)
| openai, anthropic, gemini, ollama, lmstudio, openrouter, groq, mistral, deepseek, xai, huggingface, fireworks, together, cloudflare, nvidia, azure_openai, aws_bedrock, minimax, kimi, alibaba (qwen), vllm, pollinations | Lobe Icons |
| llamacpp | Jan app asset (brand mark used for llama.cpp) |
| jan | Jan official app icon |
+| osaurus | [Osaurus official logo](https://github.com/osaurus-ai/osaurus/blob/main/App/osaurus/Assets.xcassets/osaurus-logo.imageset/osaurus-logo-black.svg) (`osaurus-ai/osaurus`, MIT) |
| localai | LocalAI official logo (`mudler/LocalAI`) |
| gpt4all | Simple mark (not an official GPT4All brand asset) |
| local_openai_proxy | WebBrain-authored generic proxy mark |
diff --git a/src/firefox/icons/providers/osaurus.svg b/src/firefox/icons/providers/osaurus.svg
new file mode 100644
index 000000000..698845466
--- /dev/null
+++ b/src/firefox/icons/providers/osaurus.svg
@@ -0,0 +1,24 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 200 200" fill="none" preserveAspectRatio="xMidYMid meet">
+ <defs>
+ <style>
+ .st0 {
+ fill: none;
+ stroke: #000000;
+ stroke-miterlimit: 10;
+ stroke-width: 16px;
+ }
+ </style>
+ </defs>
+ <g>
+ <g>
+ <line class="st0" x1="98.2" y1="191.2" x2="98.2" y2="166" />
+ <path class="st0" d="M61.6,164.1s-21.5,10.4-49.2-17.1v-.2c39.1-.2,60.1-35.9,60.1-35.9" />
+ </g>
+ <path
+ class="st0"
+ d="M61.6,191.2v-35.8c0-42.8,16.7-63.6,58.1-63.6v-43.9s-.5-29.4,34.1-29.7,34.4,15,34.4,33.7c0,18.8-9.4,27.4-22.5,32.1v29.8c0,14.2-6.2,27.8-16.9,37.2l-17.1,14.9v25.2"
+ />
+ </g>
+ <circle cx="155.8" cy="51.4" r="7.8" fill="#000000" stroke="none" />
+</svg>
diff --git a/src/firefox/src/background.js b/src/firefox/src/background.js
index 13e821364..fd4d09dfc 100644
--- a/src/firefox/src/background.js
+++ b/src/firefox/src/background.js
@@ -3509,7 +3509,9 @@ async function handleMessage(msg, sender) {
}
case 'list_provider_models': {
- return await providerManager.listProviderModels(msg.providerId);
+ return await providerManager.listProviderModels(msg.providerId, {
+ detectServerIdentity: msg.detectServerIdentity === true,
+ });
}
case 'list_ollama_models': {
diff --git a/src/firefox/src/providers/manager.js b/src/firefox/src/providers/manager.js
index 5a92916e7..8a8e096d9 100644
--- a/src/firefox/src/providers/manager.js
+++ b/src/firefox/src/providers/manager.js
@@ -53,7 +53,7 @@ import {
const WEBBRAIN_CLOUD_PROVIDER_ID = 'webbrain_cloud';
const WEBBRAIN_CLOUD_PROVIDER_LABEL = 'WebBrain Compass';
const DUPLICATE_PROVIDER_SUFFIX = '__duplicate';
-const LOCAL_MODEL_LIST_PROVIDER_IDS = ['llamacpp', 'ollama', 'lmstudio', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth'];
+const LOCAL_MODEL_LIST_PROVIDER_IDS = ['llamacpp', 'ollama', 'lmstudio', 'osaurus', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth'];
const WEBBRAIN_CLOUD_CONTEXT_WINDOW = 1000000;
const WEBBRAIN_CLOUD_LEGACY_CONTEXT_WINDOW = 256000;
const WEBBRAIN_DEVICE_GUID_KEY = 'webbrainDeviceGuid';
@@ -461,6 +461,20 @@ export class ProviderManager {
visionDetection: null,
enabled: true,
},
+ osaurus: {
+ type: 'openai',
+ category: 'local',
+ label: 'Osaurus (Local)',
+ providerName: 'osaurus',
+ requiresModel: true,
+ baseUrl: 'http://127.0.0.1:1337/v1',
+ model: '',
+ contextWindow: 16384,
+ apiKey: '',
+ supportsAskStreaming: true,
+ supportsVision: false,
+ enabled: true,
+ },
jan: {
type: 'openai',
category: 'local',
@@ -1119,7 +1133,7 @@ export class ProviderManager {
static categoryFor(id, config) {
if (config && config.category) return config.category;
if (config?.type === 'llamacpp') return 'local';
- if (['llamacpp', 'ollama', 'lmstudio', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth'].includes(id)) return 'local';
+ if (['llamacpp', 'ollama', 'lmstudio', 'osaurus', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth'].includes(id)) return 'local';
if (ROUTER_PROVIDER_IDS.includes(id)) return 'router';
return 'cloud';
}
@@ -1840,11 +1854,11 @@ export class ProviderManager {
/**
* Fetch selectable models for local providers. Ollama uses its native
- * /api/tags endpoint; llama.cpp, LM Studio, Jan, vLLM, SGLang, LocalAI,
+ * /api/tags endpoint; llama.cpp, LM Studio, Osaurus, Jan, vLLM, SGLang, LocalAI,
* GPT4All, Unsloth Studio, and generic local proxies use
* OpenAI-compatible /v1/models.
*/
- async listProviderModels(id) {
+ async listProviderModels(id, { detectServerIdentity = false } = {}) {
const provider = this.providers.get(id);
if (!provider) return { ok: false, error: 'Provider not found' };
const definitionId = this._providerDefinitionId(id, provider.config);
@@ -1916,6 +1930,13 @@ export class ProviderManager {
continue;
}
const data = await res.json();
+ // Jan and Osaurus share port 1337. Successful OpenAI model discovery
+ // alone cannot identify which server answered an unconfigured card.
+ if (detectServerIdentity && provider.config.configured !== true
+ && ['jan', 'osaurus'].includes(definitionId)) {
+ const isOsaurus = await this._isOsaurusDiscoveryServer(data, candidate.requestBaseUrl, headers);
+ if (isOsaurus !== (definitionId === 'osaurus')) return { ok: true, models: [] };
+ }
const models = this._extractModelIds(definitionId, data);
const result = { ok: true, models };
if (candidate.configBaseUrl !== observedBaseUrl) {
@@ -1943,6 +1964,28 @@ export class ProviderManager {
return firstFailure || { ok: false, error: 'Failed to load models' };
}
+ async _isOsaurusDiscoveryServer(data, requestBaseUrl, headers) {
+ if (Array.isArray(data?.data) && data.data.some(model =>
+ String(model?.owned_by || '').trim().toLowerCase() === 'osaurus'
+ )) return true;
+
+ // A cloud-only Osaurus catalog may contain only upstream owners. Its
+ // public root greeting identifies the server without depending on models.
+ const root = requestBaseUrl.replace(/\/v1$/i, '');
+ const controller = new AbortController();
+ const timer = setTimeout(() => controller.abort(), 1500);
+ try {
+ const response = await fetch(`${root}/`, {
+ method: 'GET', headers, signal: controller.signal,
+ });
+ return response.ok && (await response.text()).trim().startsWith('Osaurus Server is running!');
+ } catch {
+ return false;
+ } finally {
+ clearTimeout(timer);
+ }
+ }
+
async listOllamaModels(id) {
return this.listProviderModels(id);
}
diff --git a/src/firefox/src/ui/provider-icons.js b/src/firefox/src/ui/provider-icons.js
index 7640cc6f6..96a4bc3d5 100644
--- a/src/firefox/src/ui/provider-icons.js
+++ b/src/firefox/src/ui/provider-icons.js
@@ -15,6 +15,7 @@ export const PROVIDER_ICON_FILES = {
llamacpp: 'llamacpp.svg',
ollama: 'ollama.svg',
lmstudio: 'lmstudio.svg',
+ osaurus: 'osaurus.svg',
jan: 'jan.png',
vllm: 'vllm.svg',
sglang: 'sglang.png',
@@ -82,6 +83,7 @@ export const PROVIDER_SHORT_LABELS = {
llamacpp: 'llama.cpp',
ollama: 'Ollama',
lmstudio: 'LM Studio',
+ osaurus: 'Osaurus',
jan: 'Jan',
vllm: 'vLLM',
sglang: 'SGLang',
diff --git a/src/firefox/src/ui/settings.js b/src/firefox/src/ui/settings.js
index 1ad38990f..c52733555 100644
--- a/src/firefox/src/ui/settings.js
+++ b/src/firefox/src/ui/settings.js
@@ -2531,6 +2531,16 @@ function renderProviders() {
PROMPT_TIER_FIELD,
],
},
+ osaurus: {
+ fields: [
+ { key: 'baseUrl', labelKey: 'st.provider.field.server_url', type: 'text', placeholder: 'http://127.0.0.1:1337/v1' },
+ OPTIONAL_LOCAL_API_KEY_FIELD,
+ { key: 'model', labelKey: 'st.provider.field.model', type: 'text', placeholder: 'llama-3.2-3b-instruct' },
+ CONTEXT_WINDOW_FIELD,
+ { key: 'supportsVision', labelKey: 'st.provider.field.supports_vision', type: 'checkbox' },
+ PROMPT_TIER_FIELD,
+ ],
+ },
jan: {
fields: [
{ key: 'baseUrl', labelKey: 'st.provider.field.server_url', type: 'text', placeholder: 'http://localhost:1337/v1' },
@@ -2957,7 +2967,7 @@ function renderProviders() {
</div>
`;
} else {
- const localModelProviders = ['llamacpp', 'ollama', 'lmstudio', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth'];
+ const localModelProviders = ['llamacpp', 'ollama', 'lmstudio', 'osaurus', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth'];
const canLoadModels = localModelProviders.includes(definitionId) && field.key === 'model';
const listAttr = canLoadModels ? `list="models-${id}"` : '';
const datalistHTML = canLoadModels ? `<datalist id="models-${id}"></datalist>` : '';
diff --git a/src/firefox/src/ui/sidepanel.js b/src/firefox/src/ui/sidepanel.js
index c27b46f56..77a086927 100644
--- a/src/firefox/src/ui/sidepanel.js
+++ b/src/firefox/src/ui/sidepanel.js
@@ -132,7 +132,7 @@ if (globalThis.browser?.storage?.onChanged) {
const localModels = document.getElementById('ob-local-models');
const localModelList = document.getElementById('ob-local-model-list');
const totalSteps = steps.length;
- const LOCAL_PROVIDER_ORDER = ['unsloth', 'local_openai_proxy', 'jan', 'lmstudio', 'ollama', 'llamacpp', 'vllm', 'sglang', 'localai', 'gpt4all'];
+ const LOCAL_PROVIDER_ORDER = ['unsloth', 'local_openai_proxy', 'jan', 'osaurus', 'lmstudio', 'ollama', 'llamacpp', 'vllm', 'sglang', 'localai', 'gpt4all'];
let current = 0;
let localScanStarted = false;
let localModelChoices = [];
@@ -341,7 +341,7 @@ if (globalThis.browser?.storage?.onChanged) {
// closed port fails fast. 5s just caps a slow/stalled server. (Kept
// equal to the Chrome build for parity.)
const res = await withTimeout(
- sendToBackground('list_provider_models', { providerId }),
+ sendToBackground('list_provider_models', { providerId, detectServerIdentity: true }),
5000
);
if (res?.ok && Array.isArray(res.models)) {
diff --git a/test/run.js b/test/run.js
index f3e751119..fcb65b64c 100644
--- a/test/run.js
+++ b/test/run.js
@@ -47806,7 +47806,7 @@ test('local provider API keys stay available in a collapsed advanced section', (
/const OPTIONAL_LOCAL_API_KEY_FIELD = \{[\s\S]*?key: 'apiKey',[\s\S]*?collapsed: true,[\s\S]*?\};/,
`${label}: local providers should share one optional authentication field`,
);
- for (const id of ['llamacpp', 'ollama', 'lmstudio', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all']) {
+ for (const id of ['llamacpp', 'ollama', 'lmstudio', 'osaurus', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all']) {
const start = settings.indexOf(`${id}: {`);
assert.notEqual(start, -1, `${label}: ${id} settings missing`);
const end = settings.indexOf('\n },', start);
@@ -66877,9 +66877,201 @@ test('subscription OAuth refreshes share in-flight work and retry after failures
}
});
+test('Osaurus requires a selected model and uses the local Chat Completions contract', () => {
+ for (const PM of [ProviderManagerCh, ProviderManagerFx]) {
+ const manager = new PM();
+ const defaults = manager._defaultConfigs().osaurus;
+ const empty = manager._createProvider('osaurus', defaults);
+ assert.equal(empty.name, 'osaurus');
+ assert.equal(empty.baseUrl, 'http://127.0.0.1:1337/v1');
+ assert.equal(empty.promptTier, 'mid');
+ assert.equal(empty.supportsTools, true);
+ assert.equal(empty.supportsAskStreaming, true);
+ assert.equal(empty.supportsVision, false);
+ assert.throws(() => empty.model, /model is required/);
+
+ const provider = manager._createProvider('osaurus', { ...defaults, model: 'gpt-5-local' });
+ const body = provider._buildChatCompletionsBody([{ role: 'user', content: 'hello' }], { maxTokens: 128 });
+ assert.equal(body.model, 'gpt-5-local');
+ assert.equal(body.max_tokens, 128);
+ assert.equal('max_completion_tokens' in body, false);
+ assert.equal(provider._usesResponsesApi(), false);
+ assert.equal(provider._shouldRequestStreamUsage(), false);
+ assert.equal('Authorization' in provider._headers(), false);
+ assert.equal(manager._createProvider('osaurus', { ...defaults, supportsVision: true }).supportsVision, true);
+ }
+});
+
+test('Osaurus discovers models and handles chat, streaming, tools, and access keys', async () => {
+ const originalFetch = globalThis.fetch;
+ const calls = [];
+ const toolCall = { id: 'call_osaurus', type: 'function', function: { name: 'read_page', arguments: '{}' } };
+ globalThis.fetch = async (url, options = {}) => {
+ const body = options.body ? JSON.parse(options.body) : null;
+ calls.push({ url: String(url), headers: options.headers || {}, body });
+ if (String(url).endsWith('/models')) {
+ return new Response(JSON.stringify({ object: 'list', data: [
+ { id: 'llama-3.2-3b-instruct', owned_by: 'osaurus' },
+ { id: 'foundation', owned_by: 'osaurus' },
+ ] }), { headers: { 'Content-Type': 'application/json' } });
+ }
+ if (body.stream) {
+ const chunks = [
+ { choices: [{ delta: { content: 'Hello' }, finish_reason: null }] },
+ { choices: [{ delta: { tool_calls: [{ index: 0, ...toolCall }] }, finish_reason: null }] },
+ { choices: [{ delta: {}, finish_reason: 'tool_calls' }] },
+ ];
+ return new Response(chunks.map(chunk => `data: ${JSON.stringify(chunk)}\n\n`).join('') + 'data: [DONE]\n\n', {
+ headers: { 'Content-Type': 'text/event-stream' },
+ });
+ }
+ return new Response(JSON.stringify({ choices: [{ message: { role: 'assistant', content: 'Hello', tool_calls: [toolCall] } }] }), {
+ headers: { 'Content-Type': 'application/json' },
+ });
+ };
+ try {
+ for (const PM of [ProviderManagerCh, ProviderManagerFx]) {
+ const manager = new PM();
+ const defaults = manager._defaultConfigs().osaurus;
+ manager.providers.set('osaurus', manager._createProvider('osaurus', defaults));
+ assert.deepEqual(await manager.listProviderModels('osaurus'), {
+ ok: true, models: ['foundation', 'llama-3.2-3b-instruct'],
+ });
+ assert.equal(calls.at(-1).url, `${defaults.baseUrl}/models`);
+ assert.equal('Authorization' in calls.at(-1).headers, false);
+
+ const provider = manager._createProvider('osaurus', {
+ ...defaults, model: 'llama-3.2-3b-instruct', apiKey: 'test-access-key',
+ });
+ manager.providers.set('osaurus', provider);
+ assert.equal((await manager.testProvider('osaurus')).ok, true);
+ const tools = [{ type: 'function', function: { name: 'read_page', parameters: { type: 'object', properties: {} } } }];
+ const result = await provider.chat([{ role: 'user', content: 'Read the page' }], { tools });
+ assert.equal(result.content, 'Hello');
+ assert.deepEqual(result.toolCalls, [toolCall]);
+ assert.equal(calls.at(-1).url, `${defaults.baseUrl}/chat/completions`);
+ assert.equal(calls.at(-1).headers.Authorization, 'Bearer test-access-key');
+ assert.deepEqual(calls.at(-1).body.tools, tools);
+
+ const events = [];
+ for await (const event of provider.chatStream([{ role: 'user', content: 'Hello' }], { tools })) events.push(event);
+ assert.equal(events.filter(event => event.type === 'text').map(event => event.content).join(''), 'Hello');
+ assert.ok(events.some(event => event.type === 'tool_call'));
+ assert.ok(events.some(event => event.type === 'done'));
+ assert.equal(calls.at(-1).body.stream, true);
+ assert.equal('stream_options' in calls.at(-1).body, false);
+ }
+ } finally {
+ globalThis.fetch = originalFetch;
+ }
+});
+
+test('onboarding distinguishes Osaurus and Jan on their shared port in both browsers', async () => {
+ const originalFetch = globalThis.fetch;
+ try {
+ for (const [build, PM] of [['chrome', ProviderManagerCh], ['firefox', ProviderManagerFx]]) {
+ const panel = fs.readFileSync(path.join(ROOT, `src/${build}/src/ui/sidepanel.js`), 'utf8');
+ const orderLiteral = panel.match(/const LOCAL_PROVIDER_ORDER = (\[[^;]+\]);/)[1];
+ const order = Function(`return ${orderLiteral};`)();
+ const start = panel.indexOf(' async function scanLocalModels() {');
+ const end = panel.indexOf('\n function goTo(', start);
+ assert.ok(start !== -1 && end > start, `${build}: onboarding scan missing`);
+ const scanSource = panel.slice(start, end);
+ const scan = Function('sendToBackground', 'LOCAL_PROVIDER_ORDER', `
+ let localModelChoices = [];
+ const settingsBtn = {}, skipBtn = {}, providerBody = {};
+ const providerList = { classList: { add() {} } }, localModels = providerList;
+ const t = key => key, setProviderStatus = () => {};
+ const providerSortIndex = id => {
+ const index = LOCAL_PROVIDER_ORDER.indexOf(id);
+ return index === -1 ? LOCAL_PROVIDER_ORDER.length : index;
+ };
+ const withTimeout = promise => promise;
+ const showLocalChoices = choices => { localModelChoices = choices; };
+ const showProviderFallback = () => {};
+ ${scanSource}
+ return async () => { await scanLocalModels(); return localModelChoices; };
+ `);
+
+ for (const scenario of [
+ { name: 'local Osaurus', owner: 'osaurus', greeting: '', expected: ['osaurus'], rootCalls: 0 },
+ { name: 'cloud-only Osaurus', owner: 'openai', greeting: 'Osaurus Server is running! πŸ¦•', expected: ['osaurus'], rootCalls: 2 },
+ { name: 'Jan', owner: 'jan', greeting: '', expected: ['jan'], rootCalls: 2 },
+ { name: 'unidentified server', owner: 'unknown', greeting: 'Server ready', expected: ['jan'], rootCalls: 2 },
+ { name: 'separate Jan and Osaurus ports', separatePorts: true, greeting: '', expected: ['jan', 'osaurus'], rootCalls: 1 },
+ ]) {
+ const manager = new PM();
+ const defaults = manager._defaultConfigs();
+ const providers = { jan: defaults.jan, osaurus: { ...defaults.osaurus } };
+ if (scenario.separatePorts) providers.osaurus.baseUrl = 'http://127.0.0.1:1444/v1';
+ for (const [id, config] of Object.entries(providers)) manager.providers.set(id, manager._createProvider(id, config));
+ let rootCalls = 0;
+ globalThis.fetch = async url => {
+ const parsed = new URL(url);
+ if (parsed.pathname === '/v1/models') {
+ const owner = scenario.separatePorts ? (parsed.port === '1444' ? 'osaurus' : 'jan') : scenario.owner;
+ return new Response(JSON.stringify({ object: 'list', data: [{ id: 'text-model', owned_by: owner }] }), {
+ headers: { 'Content-Type': 'application/json' },
+ });
+ }
+ assert.equal(parsed.pathname, '/');
+ rootCalls++;
+ return new Response(scenario.greeting, { status: scenario.greeting ? 200 : 404 });
+ };
+ const requests = [];
+ const send = async (action, message) => {
+ if (action === 'get_providers') return { providers, active: 'jan' };
+ assert.equal(action, 'list_provider_models');
+ requests.push(message);
+ return manager.listProviderModels(message.providerId, {
+ detectServerIdentity: message.detectServerIdentity === true,
+ });
+ };
+ const choices = await scan(send, order)();
+ assert.deepEqual(choices.map(choice => choice.providerId), scenario.expected, `${build}: ${scenario.name}`);
+ assert.ok(requests.every(request => request.detectServerIdentity === true));
+ assert.equal(rootCalls, scenario.rootCalls, `${build}: ${scenario.name} identity probes`);
+ if (scenario.expected[0] === 'osaurus') {
+ assert.equal(manager.providers.get(choices[0].providerId).supportsVision, false, `${build}: Osaurus vision default`);
+ }
+ }
+ const background = fs.readFileSync(path.join(ROOT, `src/${build}/src/background.js`), 'utf8');
+ assert.match(background, /case 'list_provider_models':[\s\S]*?detectServerIdentity: msg\.detectServerIdentity === true/);
+ }
+ } finally {
+ globalThis.fetch = originalFetch;
+ }
+});
+
+test('Osaurus identity checks preserve manual model loading and configured providers', async () => {
+ const originalFetch = globalThis.fetch;
+ try {
+ for (const PM of [ProviderManagerCh, ProviderManagerFx]) {
+ const manager = new PM();
+ const defaults = manager._defaultConfigs();
+ let rootCalls = 0;
+ globalThis.fetch = async url => {
+ if (!String(url).endsWith('/models')) rootCalls++;
+ return new Response(JSON.stringify({ data: [{ id: 'explicit-model', owned_by: 'unknown' }] }));
+ };
+ for (const id of ['jan', 'osaurus']) {
+ manager.providers.set(id, manager._createProvider(id, defaults[id]));
+ assert.deepEqual(await manager.listProviderModels(id), { ok: true, models: ['explicit-model'] });
+ manager.providers.set(id, manager._createProvider(id, { ...defaults[id], configured: true }));
+ assert.deepEqual(await manager.listProviderModels(id, { detectServerIdentity: true }), {
+ ok: true, models: ['explicit-model'],
+ });
+ }
+ assert.equal(rootCalls, 0, 'explicit provider configuration should not require server detection');
+ }
+ } finally {
+ globalThis.fetch = originalFetch;
+ }
+});
+
test('categoryFor: local family', () => {
for (const PM of [ProviderManagerCh, ProviderManagerFx]) {
- for (const id of ['llamacpp', 'ollama', 'lmstudio', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth']) {
+ for (const id of ['llamacpp', 'ollama', 'lmstudio', 'osaurus', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth']) {
assert.equal(PM.categoryFor(id, { type: id === 'llamacpp' ? 'llamacpp' : 'openai' }), 'local');
}
assert.equal(PM.categoryFor('custom_llama_cpp', { type: 'llamacpp' }), 'local');
@@ -68208,7 +68400,7 @@ test('listProviderModels sends saved API keys for auth-enabled OpenAI-compatible
try {
for (const PM of [ProviderManagerCh, ProviderManagerFx]) {
- for (const id of ['jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth']) {
+ for (const id of ['osaurus', 'jan', 'vllm', 'sglang', 'localai', 'gpt4all', 'local_openai_proxy', 'unsloth']) {
const mgr = new PM();
const config = {
...mgr._defaultConfigs()[id],
@@ -69366,7 +69558,7 @@ test('extended provider catalog is complete, mirrored, safe, and excluded-provid
['firefox', ProviderManagerFx, 'src/firefox'],
]) {
const defaults = new PM()._defaultConfigs();
- const expectedDefaultCount = label === 'chrome' ? 110 : 109;
+ const expectedDefaultCount = label === 'chrome' ? 111 : 110;
assert.equal(
Object.keys(defaults).length,
expectedDefaultCount,