# API list prices, $ per million tokens. Used ONLY to say what the same tokens would have cost
# at API rates — nothing here bills anything. Source: the claude-api skill's model table,
# cached 2026-06-24, read 2026-09-19. Cache read is 0.1× input and cache write 1.25× input
# (the skill's README), except Fable 5.1, whose cache reads are $0.25 (stated in its section).
# `match` is a substring of the model id as the transcripts write it (without "claude-").
# Re-check against https://docs.anthropic.com/en/docs/about-claude/pricing before trusting a
# number here for a purchase decision; the value ratio on the website is a monitoring figure.
# OpenAI rows: developers.openai.com/api/docs/pricing, pasted by Jacob 2026-09-23, Standard tier,
# SHORT-context column. Every Codex rollout on this Pi names gpt-6-astra (2450 of 2450 on
# 2026-09-23). Long context (the page does not state the cut-off for the 6 series; 272K for 5.5)
# is 2× input / 1.5× output and is NOT applied — so the Codex API figure is a floor.
match	vendor	in	out	cache_read	cache_write	as_of
fable-5-1	anthropic	10.00	50.00	0.25	12.50	2026-06-24
fable-5	anthropic	10.00	50.00	1.00	12.50	2026-06-24
opus-5	anthropic	5.00	25.00	0.50	6.25	2026-06-24
opus-4-8	anthropic	5.00	25.00	0.50	6.25	2026-06-24
opus-4-7	anthropic	5.00	25.00	0.50	6.25	2026-06-24
opus-4-6	anthropic	5.00	25.00	0.50	6.25	2026-06-24
sonnet-5	anthropic	2.00	10.00	0.20	2.50	2026-06-24
sonnet-4-6	anthropic	3.00	15.00	0.30	3.75	2026-06-24
haiku-4-5	anthropic	1.00	5.00	0.10	1.25	2026-06-24
gpt-6-astra	openai	10.00	50.00	1.00	12.50	2026-09-23
gpt-6-sol	openai	2.00	10.00	0.20	2.50	2026-09-23
gpt-6-luna	openai	0.10	0.50	0.01	0.125	2026-09-23
gpt-5.3-codex	openai	1.75	14.00	0.175		2026-09-23