Files
Redflag/system/technical/synthetic_quota.md
2026-03-27 21:10:01 -04:00

1.7 KiB

description, limit
description limit
Synthetic API quota status - tokens remaining, usage stats, API response 10000

Synthetic API Quota

Last Updated: 2026-03-27T21:10:01-04:00

Current Status

API Response:

{
    "subscription": {
        "limit": 335,
        "requests": 1,
        "renewsAt": "2026-03-28T06:07:46.374Z"
    },
    "search": {
        "hourly": {
            "limit": 250,
            "requests": 0,
            "renewsAt": "2026-03-28T02:10:01.375Z"
        }
    },
    "freeToolCalls": {
        "limit": 1250,
        "requests": 410,
        "renewsAt": "2026-03-28T14:15:11.539Z"
    }
}

Usage Guidelines

Indicator Meaning Action
🟢 Green <33% usage Plenty available
🟡 Yellow 33-66% usage Monitor closely
🔴 Red >66% usage Conserve, notify Casey

Pre-Flight Check

Before starting longform tasks:

  • Check quota via synu or API response above
  • Ensure 🟢 green status before multi-agent operations
  • For 🔴 red status — notify Casey, use budget models only, or defer

Budget-Optimized Models (0.5x cost):

  • nvidia/nemotron-493b — Fast, capable, half price
  • z-ai/glm-4.7-flash — Quick inference, economical
  • Qwen/Qwen3-Coder — Coding tasks, cost-efficient

Rule of thumb: If I'm planning parallel subagents or research-heavy sessions, I check quota first. Better to ask permission than run dry mid-flight.

Quick Check

# Check quota via API directly
curl https://api.synthetic.new/v2/quotas \
  -H "Authorization: Bearer ${SYNTHETIC_API_KEY}"

# Or use synu for visual bars
synu

History

Timestamp Status Notes
2026-03-27 21:10 checked Auto-check