I decoded 23,515 local SQLite billing records on Google AI Ultra ($199.99/mo): Found ~49.5M uncached tokens from 0% KV cache drops, silent "-tiered" backend model IDs, and a Google One 30TB refund catch-22

After noticing rapid allowance drain and erratic reasoning starting Nov 15 on a $199.99/mo Google One AI Ultra (20x) plan, I decoded the raw gen_metadata protobuf billing records across all 373 local SQLite conversation DBs (~/.gemini/antigravity/conversations/*.db, 23,515 model calls).

Three verifiable findings stood out in Google’s own protobuf usage records:

  1. Server-side 0% KV Cache bug (~49.5M uncached tokens overbilled): On consecutive same-model turns <25s apart with zero compaction, Gemini 4 repeatedly dropped cached_input_tokens to 0 (e.g., 945k uncached tokens billed across two turns 25s apart on Carbon High; 3.26M uncached tokens across 150s on Orchestrator). Meanwhile, hosted Claude Opus/Sonnet 5.5 inside the exact same Antigravity client had 0 (0.0%) zero-cache failures across 623 strict follow-ups.

  2. Undisclosed “-tiered” & worker backend model IDs: Comparing field_21 (UI display label) vs field_19 (backend model string) shows Gemini 4 Hydrogen (High) silently switched on Nov 17 to “gemini-4-hydrogen-tiered-high” (with >60% lower mean thinking tokens), while 6,699 “Gemini 4 Hydrogen (Orchestrator)” calls ran on “gemini-flash-worker-agy”. Argon (Medium) calls returning thinking_tokens = 0 also jumped from 0.3% (Nov 12-13) to 64.9% (Nov 19).

  3. PSA on Google One’s Bundled 30 TB Billing Catch-22: Because Google bundles developer AI Ultra compute inside Google One storage plans, standard Tier-1 courtesy refunds immediately cancel the entire membership and revoke your 30 TB Drive/Gmail storage (to the Tier-1 rep’s credit, they warned me before clicking it—otherwise my 3.5 TB Drive & Gmail would have been frozen immediately). Only after declining cancellation and uploading raw SQLite protobuf traces + server ResponseIDs did Tier-2 acknowledge an option to refund the last plan charge without cancelling storage (still pending manual review with no SLA).

Full write-up, redacted support transcript, and zero-dependency Python self-audit script so any Pro/Ultra subscriber can check their own local SQLite DBs in 5 seconds:

v1.1 Technical Addendum (Deduplicated Telemetry + Post-Hotfix Reproducible Bugs):

Sharing an updated v1.1 audit (audit_antigravity_sqlite.py now enforces strict ResponseID deduplication across forked/Swarm DBs, requires real per-step transcript timestamps on both turns, and verifies usage.3 == usage.9 + usage.10 with 24,319/24,319 exact matches):

  1. Deduplicated Nov 15–20 KV Cache Summary: 50,650,659 overbilled uncached tokens across 397 full 0%-cache wipes within <=300s (28,028,445 within <=45s), vs 0 / 648 (0.00%) for hosted Claude in the same client. Clean <30s single-model traces with zero intervening errors or timeouts:
    • d4eb10d7 (gemini-4-carbon, +28s): lEX-aqj2IpWo**** (327,361 cached) → r0X-aqPeJpj8**** (350,781 uncached, 0 cached) → yUX-auqiBYSW**** (350,782 cached)
    • 14e9dadb (gemini-4-argon, +24s): wHj9aqnNIdHv**** (491,231 cached) → 2Hj9aujEI5z9**** (502,806 uncached, 0 cached) → Mnr9avqCLubs**** (502,807 cached after 5m46s)
  2. Post-Hotfix Reproducible Leaks Still Active on Nov 19–20 (925 turns):
    • Client Harness Bug: Async SYSTEM_MESSAGE wakeups truncate gemini-flash-worker-agy cached prefix back to ~25K–27K within 5s–84s (51.1% in 486-turn session 80ccfd72; 100% 3/3 in 611de086).
    • Orchestrator (Neon) Cache Thrashing: 71 planner/worker switches incurred 15.55M uncached input (90.3% planner miss rate after >300s worker execution), causing Orchestrator to consume 2.58x more uncached input than straight Carbon Medium.
    • 180s Server Eviction Cliff + Sub-30s Wipes: Continuous Carbon Medium stayed <3% miss under 180s, jumped to 46.2% at 180–240s and 92.9% past 300s, plus sudden 27s resets (611de086 NH8Aa8ixHM-Z****: 323,074 cached → 325,737 uncached / 0 cached).
  3. Latency Impact: For >=100K contexts, 0%-cache wipes increased median TTFT (d11) 2.57x (5.75s → 14.76s).

Full v1.1 script and updated breakdown: GitHub - fatrabbit29/antigravity-ultra-quota-audit: Read-only Python auditor & 23,515-call SQLite protobuf analysis of Google One AI Ultra (.99/mo) KV cache drops, thinking_tokens=0 spikes, and -tiered model routing in Antigravity · GitHub