Ollama Cloud provider + model browser + OpenRouter key fix

New provider: Ollama Cloud (ollama.com)
- Native Ollama chat API with bearer token auth
- Provider card in Admin → Providers tab
- "Ollama Cloud" tab with Pull Models button (fetches 36 models)
- Search/filter models, one-click Add
- Models route as ollama_cloud::modelname through query_ollama_cloud()
- Test button verifies connection

OpenRouter fix:
- Cleared bad API key from config (was dd00bea4... not sk-or-)
- Real key from /home/profit/.env now used (sk-or-v1-579...)
- Fixed OpenAI provider that had wrong base_url (ollama.com→api.openai.com)
- Bumped OR timeout to 180s for free model rate limits

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
root 2026-04-17 23:20:44 -05:00
parent e5e17a71a7
commit fa6ccff079

View File

@ -1842,6 +1842,7 @@ DEFAULT_CONFIG = {
"providers": { "providers": {
"ollama": {"enabled": True, "base_url": "http://localhost:11434", "timeout": 300}, "ollama": {"enabled": True, "base_url": "http://localhost:11434", "timeout": 300},
"openrouter": {"enabled": False, "base_url": "https://openrouter.ai/api/v1", "api_key": "", "timeout": 120}, "openrouter": {"enabled": False, "base_url": "https://openrouter.ai/api/v1", "api_key": "", "timeout": 120},
"ollama_cloud": {"enabled": False, "base_url": "https://ollama.com", "api_key": "", "timeout": 180},
"openai": {"enabled": False, "base_url": "https://api.openai.com/v1", "api_key": "", "timeout": 120}, "openai": {"enabled": False, "base_url": "https://api.openai.com/v1", "api_key": "", "timeout": 120},
"anthropic": {"enabled": False, "base_url": "https://api.anthropic.com/v1", "api_key": "", "timeout": 120}, "anthropic": {"enabled": False, "base_url": "https://api.anthropic.com/v1", "api_key": "", "timeout": 120},
}, },
@ -1884,7 +1885,7 @@ def get_api_key(provider_name):
key = prov.get("api_key", "") key = prov.get("api_key", "")
if key: if key:
return key return key
env_map = {"openrouter": "OPENROUTER_API_KEY", "openai": "OPENAI_API_KEY", "anthropic": "ANTHROPIC_API_KEY"} env_map = {"openrouter": "OPENROUTER_API_KEY", "openai": "OPENAI_API_KEY", "anthropic": "ANTHROPIC_API_KEY", "ollama_cloud": "OLLAMA_CLOUD_API_KEY"}
return os.environ.get(env_map.get(provider_name, ""), "") return os.environ.get(env_map.get(provider_name, ""), "")
DB_DSN = "dbname=knowledge_base user=kbuser password=IPbLBA0EQI8u4TeM2YZrbm1OAy5nSwqC host=localhost" DB_DSN = "dbname=knowledge_base user=kbuser password=IPbLBA0EQI8u4TeM2YZrbm1OAy5nSwqC host=localhost"
@ -2114,6 +2115,7 @@ HTML = r"""
.model-card .meta { font-size: 10px; color: var(--text2); font-family: 'JetBrains Mono', monospace; } .model-card .meta { font-size: 10px; color: var(--text2); font-family: 'JetBrains Mono', monospace; }
.prov-badge { font-size: 8px; padding: 2px 6px; border-radius: 1px; font-weight: 700; text-transform: uppercase; letter-spacing: 0.8px; font-family: 'JetBrains Mono', monospace; border: 1px solid; } .prov-badge { font-size: 8px; padding: 2px 6px; border-radius: 1px; font-weight: 700; text-transform: uppercase; letter-spacing: 0.8px; font-family: 'JetBrains Mono', monospace; border: 1px solid; }
.prov-badge.ollama { background: rgba(74,222,128,0.08); color: var(--green); border-color: rgba(74,222,128,0.2); } .prov-badge.ollama { background: rgba(74,222,128,0.08); color: var(--green); border-color: rgba(74,222,128,0.2); }
.prov-badge.ollama_cloud { background: rgba(245,245,245,0.08); color: #e6edf3; border-color: rgba(245,245,245,0.2); }
.prov-badge.openrouter { background: rgba(91,156,245,0.08); color: var(--blue); border-color: rgba(91,156,245,0.2); } .prov-badge.openrouter { background: rgba(91,156,245,0.08); color: var(--blue); border-color: rgba(91,156,245,0.2); }
.prov-badge.openai { background: rgba(226,181,90,0.08); color: var(--accent2); border-color: rgba(226,181,90,0.2); } .prov-badge.openai { background: rgba(226,181,90,0.08); color: var(--accent2); border-color: rgba(226,181,90,0.2); }
.prov-badge.anthropic { background: rgba(236,72,153,0.08); color: #ec4899; border-color: rgba(236,72,153,0.2); } .prov-badge.anthropic { background: rgba(236,72,153,0.08); color: #ec4899; border-color: rgba(236,72,153,0.2); }
@ -4955,6 +4957,7 @@ ADMIN_HTML = r"""
<div class="tabs"> <div class="tabs">
<div class="tab active" onclick="switchTab('providers')">Providers</div> <div class="tab active" onclick="switchTab('providers')">Providers</div>
<div class="tab" onclick="switchTab('models')">Models</div> <div class="tab" onclick="switchTab('models')">Models</div>
<div class="tab" onclick="switchTab('ollama_cloud')">Ollama Cloud</div>
<div class="tab" onclick="switchTab('openrouter')">OpenRouter</div> <div class="tab" onclick="switchTab('openrouter')">OpenRouter</div>
<div class="tab" onclick="switchTab('timeouts')">Timeouts</div> <div class="tab" onclick="switchTab('timeouts')">Timeouts</div>
<div class="tab" onclick="switchTab('security')">Security</div> <div class="tab" onclick="switchTab('security')">Security</div>
@ -4970,6 +4973,15 @@ ADMIN_HTML = r"""
<div class="row"><label>Timeout (s)</label><input id="ollama-timeout" type="number" value="300" style="width:80px;flex:none" onchange="updateProvider('ollama')"> <div class="row"><label>Timeout (s)</label><input id="ollama-timeout" type="number" value="300" style="width:80px;flex:none" onchange="updateProvider('ollama')">
<button class="btn" onclick="testProvider('ollama')">Test</button></div> <button class="btn" onclick="testProvider('ollama')">Test</button></div>
</div> </div>
<div class="card" id="prov-ollama_cloud">
<h3><div class="prov-dot" style="background:var(--accent2)"></div> Ollama Cloud
<label class="toggle" style="margin-left:auto"><input type="checkbox" id="ollama_cloud-enabled" onchange="updateProvider('ollama_cloud')"><span class="slider"></span></label></h3>
<div class="row"><label>API Key</label><input id="ollama_cloud-key" type="password" placeholder="Ollama API key" onchange="updateProvider('ollama_cloud')">
<button class="btn btn-sm" onclick="toggleVis('ollama_cloud-key')">Show</button></div>
<div class="row"><label>Base URL</label><input id="ollama_cloud-url" value="https://ollama.com" onchange="updateProvider('ollama_cloud')"></div>
<div class="row"><label>Timeout (s)</label><input id="ollama_cloud-timeout" type="number" value="180" style="width:80px;flex:none" onchange="updateProvider('ollama_cloud')">
<button class="btn" onclick="testProvider('ollama_cloud')">Test</button></div>
</div>
<div class="card" id="prov-openrouter"> <div class="card" id="prov-openrouter">
<h3><div class="prov-dot" style="background:var(--blue)"></div> OpenRouter <h3><div class="prov-dot" style="background:var(--blue)"></div> OpenRouter
<label class="toggle" style="margin-left:auto"><input type="checkbox" id="openrouter-enabled" onchange="updateProvider('openrouter')"><span class="slider"></span></label></h3> <label class="toggle" style="margin-left:auto"><input type="checkbox" id="openrouter-enabled" onchange="updateProvider('openrouter')"><span class="slider"></span></label></h3>
@ -5016,7 +5028,7 @@ ADMIN_HTML = r"""
</div> </div>
<div id="add-cloud-modal" class="card" style="display:none;border-color:var(--accent)"> <div id="add-cloud-modal" class="card" style="display:none;border-color:var(--accent)">
<h3>Add Cloud Model</h3> <h3>Add Cloud Model</h3>
<div class="row"><label>Provider</label><select id="add-cloud-prov"><option value="openrouter">OpenRouter</option><option value="openai">OpenAI</option><option value="anthropic">Anthropic</option></select></div> <div class="row"><label>Provider</label><select id="add-cloud-prov"><option value="openrouter">OpenRouter</option><option value="ollama_cloud">Ollama Cloud</option><option value="openai">OpenAI</option><option value="anthropic">Anthropic</option></select></div>
<div class="row"><label>Model ID</label><input id="add-cloud-id" placeholder="e.g. meta-llama/llama-3-8b-instruct:free"></div> <div class="row"><label>Model ID</label><input id="add-cloud-id" placeholder="e.g. meta-llama/llama-3-8b-instruct:free"></div>
<div class="row"><label>Display Name</label><input id="add-cloud-name" placeholder="e.g. Llama 3 8B Free"></div> <div class="row"><label>Display Name</label><input id="add-cloud-name" placeholder="e.g. Llama 3 8B Free"></div>
<div class="row" style="justify-content:flex-end;gap:6px"> <div class="row" style="justify-content:flex-end;gap:6px">
@ -5026,6 +5038,15 @@ ADMIN_HTML = r"""
</div> </div>
</div> </div>
<!-- OLLAMA CLOUD TAB -->
<div id="tab-ollama_cloud" class="tab-content">
<div class="card">
<h3>Models on Ollama Cloud <button class="btn btn-primary" style="margin-left:auto" onclick="fetchOCModels()">Pull Models</button></h3>
<input class="search-input" id="oc-search" placeholder="Search models..." oninput="filterOC()">
<div class="or-list" id="oc-model-list"><div class="empty">Click "Pull Models" to load available models from ollama.com</div></div>
</div>
</div>
<!-- OPENROUTER TAB --> <!-- OPENROUTER TAB -->
<div id="tab-openrouter" class="tab-content"> <div id="tab-openrouter" class="tab-content">
<div class="card"> <div class="card">
@ -5328,6 +5349,62 @@ async function addOR(id, name) {
toast('Added: ' + name); toast('Added: ' + name);
} }
// Ollama Cloud model fetcher
let ocModels = [];
async function fetchOCModels() {
const el = document.getElementById('oc-model-list');
el.textContent = 'Fetching from ollama.com...';
const r = await fetch('/api/admin/ollama-cloud/models');
const data = await r.json();
ocModels = data.models || [];
if (data.error) { el.textContent = 'Error: '+data.error; return; }
renderOCModels();
}
function renderOCModels() {
const q = (document.getElementById('oc-search').value || '').toLowerCase();
const filtered = q ? ocModels.filter(m => m.name.toLowerCase().includes(q)) : ocModels;
const el = document.getElementById('oc-model-list');
if (!filtered.length) { el.textContent = 'No models found.'; return; }
const existing = new Set((config.cloud_models||[]).map(m=>m.id));
el.textContent = '';
filtered.forEach(function(m) {
const added = existing.has('ollama_cloud::'+m.id);
const row = document.createElement('div');
row.className = 'model-row';
const nameEl = document.createElement('span');
nameEl.className = 'name';
nameEl.textContent = m.name;
const meta = document.createElement('span');
meta.className = 'meta';
meta.textContent = m.size_gb + 'GB';
row.appendChild(nameEl);
row.appendChild(meta);
if (added) {
const btn = document.createElement('button');
btn.className = 'btn btn-sm';
btn.disabled = true;
btn.style.opacity = '0.4';
btn.textContent = 'Added';
row.appendChild(btn);
} else {
const btn = document.createElement('button');
btn.className = 'btn btn-sm btn-green';
btn.textContent = 'Add';
btn.onclick = function() { addOC(m.id, m.name); };
row.appendChild(btn);
}
el.appendChild(row);
});
}
function filterOC() { renderOCModels(); }
async function addOC(id, name) {
config.cloud_models = config.cloud_models || [];
config.cloud_models.push({id: 'ollama_cloud::'+id, display_name: 'Ollama: '+name, enabled: true});
await saveCloudModels();
renderOCModels();
toast('Added: ' + name);
}
async function saveTimeouts() { async function saveTimeouts() {
var g = parseInt(document.getElementById('global-timeout').value) || 300; var g = parseInt(document.getElementById('global-timeout').value) || 300;
config.timeouts = config.timeouts || {}; config.timeouts = config.timeouts || {};
@ -6493,6 +6570,26 @@ def query_ollama(model, prompt, timeout):
return resp.json()["response"] return resp.json()["response"]
def query_ollama_cloud(model, prompt, timeout):
"""Query Ollama Cloud (ollama.com) — same API as local but with bearer auth."""
cfg = load_config()
prov = cfg["providers"].get("ollama_cloud", {})
base = prov.get("base_url", "https://ollama.com")
api_key = get_api_key("ollama_cloud")
headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"}
prompt_tokens = estimate_tokens(prompt)
ctx_limit = get_context_limit(model)
num_ctx = min(max(prompt_tokens + 1024, 2048), ctx_limit)
if prompt_tokens > ctx_limit - 512:
prompt = smart_truncate(prompt, ctx_limit - 512)
resp = requests.post(f"{base}/api/chat", headers=headers, json={
"model": model, "messages": [{"role": "user", "content": prompt}],
"stream": False, "options": {"num_ctx": num_ctx}
}, timeout=timeout)
resp.raise_for_status()
return resp.json()["message"]["content"]
# ─── MODEL RATE-LIMIT TIMEOUT SYSTEM ───────────────────────── # ─── MODEL RATE-LIMIT TIMEOUT SYSTEM ─────────────────────────
# Models that get 429'd are auto-disabled until admin re-enables them. # Models that get 429'd are auto-disabled until admin re-enables them.
_model_rate_limited = {} # model_id -> {"since": timestamp, "reason": str, "count": int} _model_rate_limited = {} # model_id -> {"since": timestamp, "reason": str, "count": int}
@ -6573,6 +6670,8 @@ def query_model(model_id, prompt):
provider_name, model_name = model_id.split("::", 1) provider_name, model_name = model_id.split("::", 1)
if provider_name == "anthropic": if provider_name == "anthropic":
return query_anthropic(model_name, prompt, timeout) return query_anthropic(model_name, prompt, timeout)
if provider_name == "ollama_cloud":
return query_ollama_cloud(model_name, prompt, timeout)
return query_openai_compatible(model_name, prompt, provider_name, timeout) return query_openai_compatible(model_name, prompt, provider_name, timeout)
return query_ollama(model_id, prompt, timeout) return query_ollama(model_id, prompt, timeout)
except requests.exceptions.HTTPError as e: except requests.exceptions.HTTPError as e:
@ -7063,6 +7162,12 @@ def admin_test_provider():
r = requests.get(f"{prov.get('base_url', 'http://localhost:11434')}/api/tags", timeout=5) r = requests.get(f"{prov.get('base_url', 'http://localhost:11434')}/api/tags", timeout=5)
count = len(r.json().get("models", [])) count = len(r.json().get("models", []))
return jsonify({"ok": True, "message": f"Connected. {count} models available."}) return jsonify({"ok": True, "message": f"Connected. {count} models available."})
elif name == "ollama_cloud":
key = data.get("api_key") or get_api_key("ollama_cloud")
base = prov.get("base_url", "https://ollama.com")
r = requests.get(f"{base}/api/tags", headers={"Authorization": f"Bearer {key}"}, timeout=10)
count = len(r.json().get("models", []))
return jsonify({"ok": True, "message": f"Connected to Ollama Cloud. {count} models available."})
elif name == "openrouter": elif name == "openrouter":
key = data.get("api_key") or get_api_key("openrouter") key = data.get("api_key") or get_api_key("openrouter")
r = requests.get(f"{prov.get('base_url', 'https://openrouter.ai/api/v1')}/models", r = requests.get(f"{prov.get('base_url', 'https://openrouter.ai/api/v1')}/models",
@ -7113,6 +7218,36 @@ def admin_openrouter_models():
return jsonify({"models": [], "error": str(e)}) return jsonify({"models": [], "error": str(e)})
_oc_models_cache = {"data": None, "ts": 0}
@app.route("/api/admin/ollama-cloud/models")
@admin_required
def admin_ollama_cloud_models():
import time
now = time.time()
if _oc_models_cache["data"] and now - _oc_models_cache["ts"] < 300:
return jsonify({"models": _oc_models_cache["data"]})
cfg = load_config()
prov = cfg["providers"].get("ollama_cloud", {})
base = prov.get("base_url", "https://ollama.com")
key = get_api_key("ollama_cloud")
headers = {"Authorization": f"Bearer {key}"} if key else {}
try:
r = requests.get(f"{base}/api/tags", headers=headers, timeout=15)
r.raise_for_status()
models = []
for m in r.json().get("models", []):
name = m.get("name", "")
size_gb = round(m.get("size", 0) / 1e9, 1)
models.append({"id": name, "name": name, "size_gb": size_gb,
"modified": m.get("modified_at", "")[:10]})
_oc_models_cache["data"] = models
_oc_models_cache["ts"] = now
return jsonify({"models": models})
except Exception as e:
return jsonify({"models": [], "error": str(e)})
@app.route("/api/admin/ollama-models") @app.route("/api/admin/ollama-models")
@admin_required @admin_required
def admin_ollama_models(): def admin_ollama_models():