Spaces:
Sleeping
Sleeping
File size: 8,339 Bytes
bdc2878 b1e4ece bdc2878 b1e4ece bdc2878 b1e4ece bdc2878 b1e4ece bdc2878 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 | """XAI rate-limits API protocol β fetch live quota data per mode."""
import asyncio
import orjson
from app.platform.errors import UpstreamError
from app.platform.logging.logger import logger
from app.platform.runtime.clock import now_ms
# ---------------------------------------------------------------------------
# Mode β request model name
# ---------------------------------------------------------------------------
# Each mode requires a separate POST request.
# The API returns a flat single-mode quota object per call.
_MODE_NAMES: dict[int, str] = {
0: "auto",
1: "fast",
2: "expert",
3: "heavy",
4: "grok-420-computer-use-sa",
}
# Default window durations used as fallback when API call fails.
_DEFAULT_WINDOW_SECS: dict[int, int] = {
0: 7_200, # auto β 2 h (super/heavy only)
1: 86_400, # fast β 24 h (basic; real value overrides for super/heavy)
2: 7_200, # expert β 2 h (super/heavy only)
3: 7_200, # heavy β 2 h (heavy-pool only)
4: 7_200, # grok_4_3 β 2 h (super/heavy only)
}
def _build_payload(mode_name: str) -> bytes:
"""Build rate-limits request payload: {"modelName": "fast"}"""
return orjson.dumps({"modelName": mode_name})
# ---------------------------------------------------------------------------
# Response parser
# ---------------------------------------------------------------------------
def parse_rate_limits(
body: dict, *, default_window_seconds: int = 72_000
) -> dict | None:
"""Parse flat rate-limits response.
Expected format::
{
"windowSizeSeconds": 72000,
"remainingQueries": 20,
"totalQueries": 20,
"lowEffortRateLimits": null,
"highEffortRateLimits": null
}
Returns a dict with keys ``remaining``, ``total``, ``window_seconds``
or ``None`` if the required fields are absent.
"""
remaining = body.get("remainingQueries")
if remaining is None:
return None
total = body.get("totalQueries")
window_secs = body.get("windowSizeSeconds")
return {
"remaining": int(remaining),
"total": int(total) if total is not None else int(remaining),
"window_seconds": int(window_secs) if window_secs else default_window_seconds,
}
# ---------------------------------------------------------------------------
# QuotaWindow builder
# ---------------------------------------------------------------------------
def _to_quota_window(data: dict, synced_at: int) -> object:
from app.control.account.models import QuotaWindow
from app.control.account.enums import QuotaSource
ws = data["window_seconds"]
return QuotaWindow(
remaining=data["remaining"],
total=data["total"],
window_seconds=ws,
reset_at=synced_at + ws * 1000, # estimated end of current window
synced_at=synced_at,
source=QuotaSource.REAL,
)
# ---------------------------------------------------------------------------
# HTTP fetch
# ---------------------------------------------------------------------------
async def _do_fetch(token: str, mode_name: str) -> dict:
"""POST the rate-limits endpoint for one mode and return parsed JSON body."""
from app.dataplane.reverse.transport.http import post_json
from app.dataplane.proxy import get_proxy_runtime
from app.control.proxy.models import ProxyFeedback, ProxyFeedbackKind
proxy = await get_proxy_runtime()
lease = await proxy.acquire()
try:
body = await post_json(
"https://grok.com/rest/rate-limits",
token,
_build_payload(mode_name),
lease=lease,
timeout_s=20.0,
)
await proxy.feedback(
lease, ProxyFeedback(kind=ProxyFeedbackKind.SUCCESS, status_code=200)
)
return body
except Exception as exc:
status = getattr(exc, "status", None) or getattr(exc, "status_code", None)
kind = _proxy_feedback_kind_for_error(exc, status=status)
await proxy.feedback(lease, ProxyFeedback(kind=kind, status_code=status))
raise
async def _fetch_one(token: str, mode_id: int) -> object | None:
"""Fetch quota window for a single mode. Returns QuotaWindow or None."""
mode_name = _MODE_NAMES.get(mode_id, "auto")
try:
body = await asyncio.wait_for(_do_fetch(token, mode_name), timeout=25.0)
except asyncio.TimeoutError:
logger.debug(
"rate-limits fetch timed out: token={}... mode={}", token[:10], mode_name
)
return None
except Exception as exc:
if is_invalid_credentials_error(exc):
raise
logger.debug(
"rate-limits fetch failed: token={}... mode={} error={}",
token[:10],
mode_name,
exc,
)
return None
data = parse_rate_limits(
body,
default_window_seconds=_DEFAULT_WINDOW_SECS.get(mode_id, 72_000),
)
if data is None:
logger.debug(
"rate-limits response missing quota fields: token={}... mode={} body={}",
token[:10],
mode_name,
body,
)
return None
return _to_quota_window(data, now_ms())
# ---------------------------------------------------------------------------
# Public API
# ---------------------------------------------------------------------------
async def fetch_all_quotas(
token: str, mode_ids: tuple[int, ...] | None = None
) -> dict[int, object] | None:
"""Fetch quota windows for the requested modes concurrently.
``mode_ids`` defaults to ``(0, 1, 2, 3, 4)``. Returns ``{mode_id: QuotaWindow}``
for every mode that responded successfully, or ``None`` if every requested
mode failed.
"""
import asyncio
requested = mode_ids or (0, 1, 2, 3, 4)
results = await asyncio.gather(
*(_fetch_one(token, mode_id) for mode_id in requested), return_exceptions=True
)
for result in results:
if isinstance(result, Exception) and is_invalid_credentials_error(result):
raise result
windows = {
mode_id: win
for mode_id, win in zip(requested, results, strict=False)
if win is not None and not isinstance(win, Exception)
}
return windows if windows else None
async def fetch_mode_quota(token: str, mode_id: int) -> object | None:
"""Fetch the quota window for a single mode. Returns QuotaWindow or None."""
return await _fetch_one(token, mode_id)
def is_invalid_credentials_body(body: str) -> bool:
"""Return whether *body* contains a Grok invalid/blocked account marker."""
text = str(body or "").lower()
return (
"invalid-credentials" in text
or "bad-credentials" in text
or "failed to look up session id" in text
or "blocked-user" in text
or "email-domain-rejected" in text
or "session not found" in text
or "account suspended" in text
or "token revoked" in text
or "token expired" in text
)
def is_invalid_credentials_error(exc: BaseException) -> bool:
"""Return whether *exc* indicates the account is invalid or blocked."""
if not isinstance(exc, UpstreamError):
return False
if exc.status not in (400, 401, 403):
return False
return is_invalid_credentials_body(str(exc.details.get("body", "") or ""))
def _proxy_feedback_kind_for_error(
exc: BaseException,
*,
status: int | None,
):
"""Map quota-fetch failures to proxy feedback without burning healthy clearance."""
from app.control.proxy.models import ProxyFeedbackKind
# Invalid or blocked accounts are account problems, not proxy problems.
if is_invalid_credentials_error(exc):
return ProxyFeedbackKind.FORBIDDEN
if status == 429:
return ProxyFeedbackKind.RATE_LIMITED
if status == 403:
return ProxyFeedbackKind.CHALLENGE
if status == 401:
return ProxyFeedbackKind.UNAUTHORIZED
if status and status >= 500:
return ProxyFeedbackKind.UPSTREAM_5XX
return ProxyFeedbackKind.TRANSPORT_ERROR
__all__ = [
"parse_rate_limits",
"fetch_all_quotas",
"fetch_mode_quota",
"is_invalid_credentials_body",
"is_invalid_credentials_error",
]
|