Plan guide

How much coding do you get from Codex Pro 500?

Available models: GPT-6 Astra / GPT-6.1 Sol / GPT-6 Luna. $500/mo buys about 45 h/week of building at p50 peak (p10–p90 20–80) — hours are guessed, so treat them as a guess. Coding-agent quality scores (/100): GPT-6 Astra (max) · Codex · 62/100 planning, GPT-6 Astra (max) · Codex · 62/100 building, GPT-6 Astra (max) · Codex · 62/100 admin. Building model cost: $7.50 per benchmark task at API prices. Catalog note: ChatGPT Pro 500 (DevDay 2026): 25x the Plus allowance plus Ultrafast mode — note Ultrafast burns the included allowance at 8x the Standard rate. Hours haircut hard from the 25x banner (Plus anchor x25 would be ~100 building h/week). Guessed — no lived reports yet.

Hours by role

p10 / p50 / p90, h/week, at the plan's own clock. Peak and off-peak differ only where the vendor publishes a night rate.

RoleBenchmark quality /100Peak p10 / p50 / p90 (h/week)Off-peak p10 / p50 / p90 (h/week)
PlanningGPT-6 Astra (max) · Codex · 62/10010 / 24 / 48—
BuildingGPT-6 Astra (max) · Codex · 62/10020 / 45 / 80—
AdminGPT-6 Astra (max) · Codex · 62/10025 / 55 / 100—

Clock: no published off-peak discount. Runs as a CLI agent. Source: guessed.

Get Codex Pro 500 →Rank it against your stack →

Current coding-agent evaluations

The ranking uses the fixed setting named in each role above. These scores do not establish subscription hours. Model availability verified 2026-11-08: vendor documentation.