What image generation actually costs, per model

Fri, 21 Aug 2026 12:00:00 GMT

One prompt, six image models, the real bill for each: 1 to 10 credits for the same 1024x1024 picture, measured against the live API with every generated image linked.

Every model Luxin can run, and what each one charges

Fri, 21 Aug 2026 12:00:00 GMT

The full generative-media catalogue with the real per-image credit debit, generated from the live production registry on every deploy.

luxin-cli 0.2.9

Mon, 17 Aug 2026 12:00:00 GMT

Discovery, batch work, and recovery — the release for agents doing real multi-image jobs.

luxin-cli 0.2.8

Wed, 12 Aug 2026 12:00:00 GMT

New command: luxin skill prints the full usage guidance from bundled package bytes, so instructions always match the CLI you are running. The installed SKILL.md is now a ~4KB discovery stub that points here; the full body is also served at https://luxin.sh/skill-core.md. · Skill freshness handshake: export LUXIN_SKILL_REVISION=

luxin-cli 0.2.7

Wed, 12 Aug 2026 12:00:00 GMT

Release (pricing): the guide's local credit quoting for xAI models moved from a startsWith("xai.grok-imagine-image") prefix test to an explicit model-id list. The prefix test would have priced any Grok Imagine model it had never heard of off the 1.x table — for the new xai.grok-imagine-image-2.0 that meant quoting 4 credits for a 10-credit create. Models outside the list now defer to the hosted registry's published debit instead of guessing. · Adds the xai.grok-imagine-image-2.0 quality×resolution credit matrix to local guide pricing (low/1k through medium/2k), matching the registry the hosted

luxin-cli 0.2.6

Sun, 09 Aug 2026 12:00:00 GMT

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 d

luxin-cli 0.2.5

Sun, 09 Aug 2026 12:00:00 GMT

Release (correctness): edit --guide selects a model again. It gated on media.input.images.required, which the hosted /v1/models summary does not publish, so on 0.2.4 every edit brief returned no_executable_model and the blocker named create. The predicate now reads supports, which the summary does publish; measured on the live 220-model catalog the two express the same 79-model set, so this drops a redundant conjunct rather than a guarantee. The blocker also names the operation it actually failed on. · Release (correctness): a plain video brief reports its aspect ratio again. selection.suggest

luxin-cli 0.2.4

Sun, 09 Aug 2026 12:00:00 GMT

Release (correctness): the guide honors an aspect ratio stated in the brief instead of dropping it (#2213, #2203, #2209). A "16:9 cinematic widescreen poster" brief previously returned a 2048x2048 square and charged 12 credits for it: selection.suggested_aspect_ratio was hard-null for every image model because the guide only ever suggested a ratio for video. The guide now extracts an explicit ratio conservatively, treats it as a required capability during model selection the way #2225 treats seed, and passes --aspect-ratio through to create, dry-run and edit. Bare "portrait" and "landscape" de

luxin-cli 0.2.3

Tue, 04 Aug 2026 12:00:00 GMT

Release (self-fund/recovery): the wall now offers a human-funding handoff instead of only a rail no external agent could walk (#2201). Both wall types — credits depleted and daily job cap reached — route through QUOTA_EXCEEDED, and the recovery envelope now carries error.recovery.human_handoff: a copy-runnable luxin claim code --json, typed no_spend, with the honest instruction to hand data.dashboard_url (and nothing else — never the token, never the raw code) to your human, who funds credits at the dashboard in about a minute. With no wallet evidence — the external default — that handoff lead

luxin-cli 0.2.2

Thu, 30 Jul 2026 12:00:00 GMT

Release (retention/self-fund): publish the human handoff from #2181. New verb luxin claim code mints a single-use, 15-minute dashboard link for the authenticated agent's human and hands back data.dashboard_url — give your human that link and they can watch your work and fund your credits there, signing in with nothing but the link (no email, no password, no provider billing setup). The mint envelope teaches the handoff in-band rather than only in docs: data.human_handoff carries share_field: "data.dashboard_url" and never_share: ["token", "code"] alongside a warning with the same sentence, so

luxin-cli 0.2.1

Tue, 28 Jul 2026 12:00:00 GMT

Release (activation/self-fund): publish the first luxin-cli payload since the 0.2.0 rename so external agents receive the credits-as-funded-boundary quota contract from #2149 — the starter preview is 50 lifetime credits with a 50-job UTC-day compatibility cap, funded identities have no daily job cap because prepaid credits and atomic credit reservations are the spend boundary, and usage quota omits daily_jobs for funded identities — plus the guide budget-basis fix from #2164 so create --guide picks hosted defaults against the same credit debit the hosted budget guard enforces (a max_estimated_

luxin-cli 0.1.72

Fri, 03 Jul 2026 12:00:00 GMT

Release (activation/self-fund): publish the quota-wall recovery contract from #2064 so error.recovery.suggested_command points directly at the browserless x402 top-up quote while error.recovery.top_up.first_command remains the no-spend payment-method inspection path. No payment caps, wallet settlement, provider spend, hosted deploys, production writes, or media generation behavior changed in this release bump.

luxin-cli 0.1.71

Fri, 03 Jul 2026 12:00:00 GMT

Release (agent UX/trust): publish the public contract guidance from #2057 clarifying that idempotency keys, recovery commands, and in-flight reservation breadcrumbs are agent-internal details during normal progress updates. No payment caps, wallet action, provider spend, hosted deploy, production write, or media generation behavior changed in this release bump.

luxin-cli 0.1.70

Thu, 25 Jun 2026 12:00:00 GMT

Release (activation/self-fund): publish the post-success self-fund handoff from #2042 so hosted create/edit successes can expose a top-level data.self_fund_next_command when quota says a top-up is recommended and the quote command is copy-runnable. This release prep does not change payment caps, wallet settlement, provider spend, hosted deploys, production writes, or media generation behavior.

luxin-cli 0.1.69

Thu, 25 Jun 2026 12:00:00 GMT

Release (activation/discovery/self-fund): publish the post-0.1.68 public CLI payload so external image-skill@latest agents receive copy-runnable self-fund commands on hosted quota walls, the public agent-skill discovery surface in CLI help/manifest output, and model discovery query/alias support for faster provider/model selection. This release prep does not change payment caps, wallet settlement, provider spend, hosted deploys, production writes, or media generation behavior.

luxin-cli 0.1.68

Thu, 18 Jun 2026 12:00:00 GMT

Release (activation/self-fund): publish the quote-auth retry handoff from #2002 so a valid unauthenticated credits quote failure preserves the exact error.recovery.after_auth retry command, plus env-token and stdin-token handoffs, after signup. The quote remains no-charge: no buy, wallet settlement, provider call, credit debit, media write, hosted deploy, or production write changed in this release bump.

luxin-cli 0.1.67

Thu, 18 Jun 2026 12:00:00 GMT

Activation/self-fund: when create --guide reaches ready_to_create and pre-wall self-fund setup is recommended, data.self_fund_next_command now aliases data.self_fund_preparation.quote_command with a pre_wall_... label so agents can open the x402 quote path before consuming remaining free allowance. Guide mode remains no-spend; buy/status/wallet settlement still require delegated spend authority.

luxin-cli 0.1.66

Thu, 18 Jun 2026 12:00:00 GMT

Release (activation/self-fund): publish ready-to-create guide data.self_fund_preparation urgency fields, so agents that still have free allowance can see setup_recommended, urgency_score, and urgency_reasons before they spend down into the quota wall. The public CLI also derives the same urgency from older hosted quota responses that only carry recommendation_reason. No payment caps, wallet action, provider spend, hosted deploy, or production write changed in this release bump.

luxin-cli 0.1.65

Wed, 17 Jun 2026 12:00:00 GMT

Release (activation/self-fund): make hosted create/edit success data.quota.top_up and data.next_actions.self_fund commands copy-runnable from the public CLI, so near-wall agents can move from a successful media response into the same zero-setup npx image-skill@latest top-up path as quota reads and quota errors. No payment caps, wallet action, provider spend, hosted deploy, or production write changed in this release bump.

luxin-cli 0.1.64

Wed, 17 Jun 2026 12:00:00 GMT

Release (activation/self-fund): publish the hosted success-surface data.next_actions.self_fund visibility payload under an unpublished package version so image-skill@latest can expose the no-spend self-fund first command whenever hosted quota reports top-up availability. recommended and urgency remain priority signals; availability controls whether the action is visible. No payment caps, auth semantics, provider routing, wallet action, provider spend, media spend, hosted deploy, or production write changed in this release bump.

luxin-cli 0.1.63

Wed, 17 Jun 2026 12:00:00 GMT

Release (activation/self-fund): make quota/payment error recovery top-up commands copy-runnable from ephemeral npx image-skill@latest runs. When hosted create/edit or payment commands return error.recovery.top_up, the public CLI now prefixes suggested_command, suggested_commands, top_up.first_command, top_up.quote_command, nested top_up.commands.*, and workflow step commands with the same zero-install handoff used by usage quota --json and create --guide.

luxin-cli 0.1.62

Wed, 17 Jun 2026 12:00:00 GMT

Release (activation/self-fund): make usage quota --json top-up handoffs copy-runnable from ephemeral npx image-skill@latest runs. Quota top_up, generated next_actions.self_fund, and create-guide embedded quota top-up commands now carry the public handoff prefix instead of bare image-skill ... commands.

luxin-cli 0.1.61

Wed, 17 Jun 2026 12:00:00 GMT

Release (activation/self-fund): add payment-attempt continuation actions. credits status --quote-id ... now repeats the exact data.next_actions.recommended_buy command for open x402/Checkout quotes, and Stripe x402 buy/status responses expose data.next_actions.recommended_settlement with exact payable-instruction paths, status/quota commands, and wallet-settlement safety flags.

luxin-cli 0.1.60

Wed, 17 Jun 2026 12:00:00 GMT

Release (activation/self-fund): add credits quote --json data.next_actions.recommended_buy, carrying the returned quote_id, a stable purchase idempotency key, copy-runnable buy/status commands, and explicit live-money payment-attempt safety fields. Quote remains no-credit and no-media; buy still requires delegated spend and verified settlement/webhook fulfillment before credits are granted.

luxin-cli 0.1.59

Wed, 17 Jun 2026 12:00:00 GMT

Release (activation/self-fund): add credits methods --json data.next_actions.recommended_quote, a ranked copy-runnable quote handoff that prefers browserless agent-settleable x402 when available and falls back to Stripe Checkout when that is the only usable live rail. The quote step is explicitly no-spend: it creates a live-money payment object but does not move money, grant credits, debit credits, call a provider, or write media.

luxin-cli 0.1.58

Wed, 17 Jun 2026 12:00:00 GMT

Release (activation/self-fund): mirror top-up urgency and the no-spend payment rail inspection command into hosted generated-asset assets show/get data.next_actions.self_fund, so agents that inspect or download created media can still discover the safe first self-fund step before ending the session. No payment caps, auth semantics, provider routing, wallet action, provider spend, media spend, hosted deploy, or production write changed in this release bump.

luxin-cli 0.1.57

Tue, 16 Jun 2026 12:00:00 GMT

Release (activation/self-fund): mirror top-up urgency and the no-spend payment rail inspection command into hosted activity list/show data.next_actions.self_fund, so agents inspecting their ledger after generated work can discover the safe first self-fund step before hitting the quota wall. No payment caps, auth semantics, provider routing, wallet action, provider spend, media spend, hosted deploy, or production write changed in this release bump.