docs(pricing): the published Gemini rates were quoting a bug we just fixed - #24
Merged
Conversation
…fixed blockrun#304 corrects google/gemini-3.5-flash from 0.50/3.00 to Google's 1.50/9.00. /v1/models and /api/pricing read models.ts live, so the moment that lands the API quotes 1.50 while every page here still says 0.50. Advertising a third of what we charge is worse than the underbilling it replaces: underbilling was invisible to the caller, a stale quote is not. llms.txt and agent.md are read by agents to budget before they call. Also fixes gemini-3.1-flash-lite in api-reference/models.md, which carried 2.5-flash-lite's 0.10/0.40 against a real 0.25/1.50. Same copy-paste-from- a-sibling shape as the bug that started this, found while checking the column convention. Raw COGS everywhere except products/intelligence/pricing.md, which quotes post-markup and now reads 1.58/9.45. The "$1 gets you" row drops from ~2M to ~635K input tokens. The clawrouter savings figure for 3.5-flash is recomputed against the baseline that row implied for itself (3.00 -> 90%), giving 70% at 9.00. That table has no single baseline across its rows, so treat 70% as arithmetic on one row, not a re-derived marketing claim.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Companion to blockrun#304, which corrects
google/gemini-3.5-flashfrom0.50/3.00to Google's1.50/9.00./v1/modelsand/api/pricingreadmodels.tslive, so the moment #304 lands the API quotes $1.50 while every page in this repo still says $0.50. Advertising a third of what we charge is a worse failure than the underbilling it replaces: the underbilling was invisible to the caller, a stale public quote is not.Changed
docs/api-reference/models.md$0.50/M | $3.00/M$1.50/M | $9.00/Mdocs/products/intelligence/overview.md$0.50/M | $3.00/M$1.50/M | $9.00/Mdocs/products/intelligence/pricing.md(post-markup)$0.53 | $3.15$1.58 | $9.45docs/products/intelligence/pricing.md($1 buys)~2M input tokens~635K input tokensdocs/products/routing/clawrouter.md(MEDIUM tier)$0.50/M in / $3.00/M out$1.50/M in / $9.00/M outdocs/products/routing/clawrouter.md(example)$3.00/M | 90%$9.00/M | 70%docs/resources/ecosystem.md$0.50–$2.00 / $3.00–$12.00$1.50–$2.00 / $9.00–$12.00Raw COGS everywhere except
products/intelligence/pricing.md, which quotes post-markup (verified: its Gemini 3.1 Pro row is$2.10/$12.60=2.00/12.00 × 1.05).Second bug, same shape
docs/api-reference/models.mdlistedgemini-3.1-flash-liteat$0.10/M | $0.40/M. That isgemini-2.5-flash-lite's rate verbatim; the real one is0.25/1.50. Found while checking which convention the column used. Fixed here, now$0.25/M | $1.50/M.On the 70%
The clawrouter savings column has no consistent baseline across its rows (
$15.00 → 80%implies 75,$0.40 → 99%implies 40,$3.00 → 90%implies 30). 70% is arithmetic against the baseline that row implied for itself, not a re-derived marketing claim. Worth a proper pass separately.Not changed, needs a product call
docs/products/intelligence/pricing.md:216still recommends Gemini 3.5 Flash for "Quick responses — Fast + cheap". At $1.50/$9.00 it is no longer the cheap option;gemini-3-flash-previewat0.50/3.00and Claude Haiku 4.5 at0.80/4.00both undercut it. Changing which model the docs recommend is a product decision, not a price sync, so it is left alone here.Verified
blockrunfull suite against this tree: 1086 tests, 99 files, all pass, includingbrand-numbers.docs.test.tswhich walks this repo'sdocs/.tsc --noEmitclean.