Merge branch 'feat/per-provider-rate-limit-per-second'

aisbf/__init__.py

@@ -55,7 +55,7 @@ from .auth.qwen import QwenOAuth2
 from .handlers import RequestHandler, RotationHandler, AutoselectHandler
 from .utils import count_messages_tokens, split_messages_into_chunks, get_max_request_tokens_for_model, get_max_completion_tokens_for_model
 
-__version__ = "0.99.95"
+__version__ = "0.99.96"
 __all__ = [
     # Config
     "config",

aisbf/config.py

@@ -112,6 +112,7 @@ class ProviderConfig(BaseModel):
     # Per-provider client rate limit: a global cap (all clients combined) on
     # requests forwarded to this provider, protecting the upstream from hammering
     # (e.g. exhausting a per-IP free-model quota). 0 = unlimited.
+    client_rate_limit_rps: int = 0   # max client requests per second to this provider
     client_rate_limit_rpm: int = 0   # max client requests per minute to this provider
     client_rate_limit_rph: int = 0   # max client requests per hour to this provider
     api_key: Optional[str] = None  # Optional API key in provider config

aisbf/handlers.py

@@ -196,24 +196,20 @@ def _provider_rl_check(bucket: str, window_seconds: int, max_requests: int) -> t
 
 
 def _provider_rl_limits(provider_config) -> tuple:
-    """Extract (rpm, rph) client rate limits from a provider config (object or dict)."""
+    """Extract (rps, rpm, rph) client rate limits from a provider config (object or dict)."""
     if provider_config is None:
-        return 0, 0
+        return 0, 0, 0
     if isinstance(provider_config, dict):
-        rpm = provider_config.get('client_rate_limit_rpm', 0)
-        rph = provider_config.get('client_rate_limit_rph', 0)
+        get = provider_config.get
     else:
-        rpm = getattr(provider_config, 'client_rate_limit_rpm', 0)
-        rph = getattr(provider_config, 'client_rate_limit_rph', 0)
-    try:
-        rpm = int(rpm or 0)
-    except (TypeError, ValueError):
-        rpm = 0
-    try:
-        rph = int(rph or 0)
-    except (TypeError, ValueError):
-        rph = 0
-    return rpm, rph
+        get = lambda k, d=0: getattr(provider_config, k, d)
+    out = []
+    for key in ('client_rate_limit_rps', 'client_rate_limit_rpm', 'client_rate_limit_rph'):
+        try:
+            out.append(int(get(key, 0) or 0))
+        except (TypeError, ValueError):
+            out.append(0)
+    return tuple(out)
 
 
 def _enforce_provider_client_rate_limit(provider_id: str, provider_config) -> None:
@@ -224,8 +220,8 @@ def _enforce_provider_client_rate_limit(provider_id: str, provider_config) -> No
     the provider's upstream quota. 0 = unlimited (the default), so providers
     without a configured limit are unaffected.
     """
-    rpm, rph = _provider_rl_limits(provider_config)
-    for window, limit, label in ((60, rpm, "per-minute"), (3600, rph, "per-hour")):
+    rps, rpm, rph = _provider_rl_limits(provider_config)
+    for window, limit, label in ((1, rps, "per-second"), (60, rpm, "per-minute"), (3600, rph, "per-hour")):
         allowed, retry_after = _provider_rl_check(f"provider:{provider_id}:{window}", window, limit)
         if not allowed:
             logging.getLogger(__name__).warning(

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "aisbf"
-version = "0.99.95"
+version = "0.99.96"
 description = "AISBF - AI Service Broker Framework || AI Should Be Free - A modular proxy server for managing multiple AI provider integrations"
 readme = "README.md"
 license = "GPL-3.0-or-later"

setup.py

@@ -106,7 +106,7 @@ class InstallCommand(_install):
 
 setup(
     name="aisbf",
-    version="0.99.95",
+    version="0.99.96",
     author="AISBF Contributors",
     author_email="stefy@nexlab.net",
     description="AISBF - AI Service Broker Framework || AI Should Be Free - A modular proxy server for managing multiple AI provider integrations",

templates/dashboard/providers.html

@@ -1810,6 +1810,10 @@ function renderProviderDetails(key) {
 
         ${subPanel(`providers:${key}:ratelimit`, window.i18n.t('providers.rate_limit_section') || 'Rate Limiting', `
         <div style="border-left: 3px solid #4a9eff; padding-left: 12px;">
+            <div class="form-group">
+                <label>${window.i18n.t('providers.client_rate_limit_rps') || 'Max client requests per second'} (0 = unlimited)</label>
+                <input type="number" min="0" value="${provider.client_rate_limit_rps || 0}" onchange="updateProvider('${key}', 'client_rate_limit_rps', this.value ? parseInt(this.value) : 0)">
+            </div>
             <div class="form-group">
                 <label>${window.i18n.t('providers.client_rate_limit_rpm') || 'Max client requests per minute'} (0 = unlimited)</label>
                 <input type="number" min="0" value="${provider.client_rate_limit_rpm || 0}" onchange="updateProvider('${key}', 'client_rate_limit_rpm', this.value ? parseInt(this.value) : 0)">

templates/dashboard/user_providers.html

@@ -1492,6 +1492,10 @@ function renderProviderDetails(key) {
 
         <div style="background: var(--bg-panel); padding: 15px; border-radius: 5px; margin: 15px 0; border-left: 3px solid #4a9eff;">
             <h4 style="margin: 0 0 15px 0; color: var(--color-link);">${window.i18n.t('providers.rate_limit_section') || 'Rate Limiting'}</h4>
+            <div class="form-group">
+                <label>${window.i18n.t('providers.client_rate_limit_rps') || 'Max client requests per second'} (0 = unlimited)</label>
+                <input type="number" min="0" value="${provider.client_rate_limit_rps || 0}" onchange="updateProvider('${key}', 'client_rate_limit_rps', this.value ? parseInt(this.value) : 0)">
+            </div>
             <div class="form-group">
                 <label>${window.i18n.t('providers.client_rate_limit_rpm') || 'Max client requests per minute'} (0 = unlimited)</label>
                 <input type="number" min="0" value="${provider.client_rate_limit_rpm || 0}" onchange="updateProvider('${key}', 'client_rate_limit_rpm', this.value ? parseInt(this.value) : 0)">