		@@ -9,2 +9,19 @@ # Changelog

		## 0.2.6 - 2026-08-09

		- Release (trust): `create --guide` now says when it cannot price the request it
		is planning, instead of presenting the model's default-shape quote as the
		planned debit. `selection.planned_debit_source` reports which of the two you
		got, and the unpriceable case carries a warning pointing at the dry-run.
		- This closes a gap that was invisible rather than wrong: the hosted API
		published per-aspect-ratio create debits only when they differed from the
		quote, so an absent field could mean either "the quote already is the debit"
		or "this server is older than your CLI". The npm package and the hosted API
		deploy independently, so that second case is permanent, not a one-off, and it
		under-quoted silently during any deploy gap or rollback. The API now publishes
		the debit for every create-capable model (107 of 107 live rows), so absence
		means exactly one thing, and the CLI now treats it that way.
		- Carries the correction to the 0.2.5 note below: the "known issue" that shipped
		in the 0.2.5 tarball describes behaviour that does not exist.

		## 0.2.5 - 2026-08-09
		@@ -23,8 +40,11 @@
		ratio.
		- Known issue (#2259): `create --guide` can quote a different credit figure than
		the `--dry-run` it hands you, because it prices the model's catalog default
		shape rather than the ratio the command actually sends. Measured against the
		live API: 4 vs 5 with no ratio stated, 4 vs 3 at 16:9. The dry-run figure is
		the authoritative one; budget from it, not from the guide. Unchanged from
		0.2.4; the fix landed server-side but not in this package.
		- Correction (added after 0.2.5 shipped): the 0.2.5 tarball carries a "known
		issue" note claiming `create --guide` can quote a different credit figure than
		the `--dry-run` it hands you. **That note is wrong and the behaviour it
		describes does not exist.** It was written from a measurement taken during a
		deploy-skew window, minutes after the fix merged and before the hosted API had
		deployed its half. Measured after both halves landed, the guide and the dry-run
		agree: 5 and 5 with no ratio stated, 3 and 3 at 16:9. #2259 is fixed and
		closed. Budgeting from the dry-run figure remains sound advice, but it is not a
		workaround for anything.

		@@ -31,0 +51,0 @@ ## 0.2.4 - 2026-08-09

+1

-1

package.json

		{
		"name": "luxin-cli",
		"version": "0.2.5",
		"version": "0.2.6",
		"description": "Image generation, image editing, video, audio, and image-to-3D for agents, with no API key to obtain first. The agent signs itself up, starts on free preview credits, and gets hosted media URLs, recoverable jobs, and cost receipts.",
		@@ -5,0 +5,0 @@ "type": "module",

bin/luxin.mjs

Sorry, the diff of this file is too big to display

luxin-cli - npm Package Compare versions

New alerts

Improved metrics

Worsened metrics