Fashion Cross-Reasoning.
Three intensities.

No. Apshan never auto-charges beyond your subscription. Top-ups are explicit purchases. You set the budget. Usage is capped at your monthly allowance, and exceeding it pauses the metered features until you choose to buy more credits.

Buy credit top-ups in flat increments. €100 for 1,000 credits. Credits apply across research and image generation. Crest and Apex unlock progressively better top-up rates.

Top-up credits never expire. Monthly included caps reset on your billing date and do not roll over. That is what keeps the subscription price predictable. If you consistently leave caps unused, downgrade. If you consistently exceed them, the credit top-up covers the surplus.

Apshan is accessed through Mistral or Claude, the two AI assistants Apshan integrates with today. It adds fashion-native intelligence to the assistant, not another chat to switch to. You keep your existing subscription. Apshan extends what your assistant can answer.

Every individual tier (Vista, Crest, Apex) runs through Mistral and Claude and reads every domain. Deep research against the database Apshan built. Image generation. Live search investigations scale by tier (core on Vista, in-depth on Crest, extensive on Apex). Each output ships with citations, reasoning, and confidence per fact. Crest and Apex also add image-generation sovereignty features.

Crest is sized for an active solo professional with 2,500 credits a month and up to in-depth live search investigations. Apex is for someone running near production volume: 4× the credits (10,000 / month), extensive live search investigations, dedicated brand fingerprint, audit-grade provenance, priority queues across all surfaces.

Not a list of search results. Apshan dispatches an AI research agent that scours the web, follows citations, cross-checks sources, and synthesizes a structured answer. Every fact ships with its source, the reasoning that found it, and a confidence score. Closer to a junior analyst's briefing than to Google. Three depth modes: Core covers most queries (a handful of sources, minutes). In-depth widens scope (more sources, longer reasoning). Extensive runs analyst-grade multi-step investigations across many sources. Deeper modes burn more compute (more sources read, more inference, more cross-checking), so you only pay for the depth you actually pick per query.

Team contracts are negotiated by scope: number of seats, brand workspaces, compliance requirements, deployment model. A single anchor would either misprice a small DNVB or undercharge a group. Industry standard at this contract size.

Yes. Annual prepayments are non-refundable but you can upgrade mid-cycle and the unused balance prorates to the new tier. Downgrades take effect at the next billing period.

Every new account starts with 100 free credits. No card required, email only. The free credits are valid for 14 days and cover deep research against our own database. Image generation and live search investigations are paid-only because they run real backend cost on every call. Beyond the trial, Vista €39/mo is the entry point.