Codex
Tracked across 45 snapshots (2025-12-09 → 2026-07-19).
Rating timeline
Dimension trajectory
Autonomy
12/20Integration
16/20Context
16/20Compliance
12/20Viability
18/20Interface
16/20Cap timeline
Notable events
- 2026-01-09Drop-12 rating
- 2026-01-12Jump+7 rating
- 2026-02-18Cap +reliability-complaints
- 2026-04-20Cap −reliability-complaints
- 2026-05-16Cap +reliability-complaints
- 2026-05-16Cap +pricing-volatility
- 2026-06-10Drop-5 rating
What would move this next
Upgrade to Validated if: (1) reliability-complaints cap cleared — 90+ days with no Codex-touching status incidents AND independent evidence that GPT-5.6-era output-trustworthiness concerns (METR reward-hacking, system-card fabrication rates) are remediated; clock resets to Jul 11, so earliest clearance ~early-Oct 2026; (2) pricing-volatility cap cleared — 12 months billing stability (earliest ~Apr 2027; July's usage-pool wave and the undated 5-hour-limit removal restart the narrative); (3) HIPAA cloud execution supported (not just local environments); (4) a hands-on UX/end-to-end assessment unlocks band 5 on the capability dims currently desk-capped at 16. Compliance moves toward 16-18 when pricing-volatility clears, provided CVE-2026-35603 is also remediated. Demotion to Detected if: billing anomalies escalate to enterprise contract disputes, reliability materially worsens past current levels, or CVE-2026-35603 is exploited in the wild while unpatched. Monitor: Codex incident trend under GPT-5.6 (uptime improved in early July — does it hold); whether/how the 5-hour limit returns and whether the shared Work+Codex usage pool is restructured; confirmation of GPT-5.6 availability for Codex cloud tasks (unconfirmed report says cloud tasks still route GPT-5.5); any CVE-2026-35603 fix (none through CLI 0.144.3 / Jul 13); a newer weekly-user figure (5M+ is six weeks stale); Ninth Circuit appeal movement; whether exec turnover continues (Simo, Heidecke, Achiam in July alone).