Claude Projects is Anthropic's cloud workspace for organizing your Claude.ai conversations. Advisor Prep Hero is a local-first desktop alternative where every Claude conversation becomes a Markdown file on your machine. Same Claude model under the hood. The split: who owns the conversation history and where it lives, which matters a great deal if you're an attorney, CPA, or consultant with confidentiality obligations.
| Feature | Claude Projects | Advisor Prep Hero |
|---|---|---|
| Where conversations live | Anthropic's servers | Your hard drive (Markdown files) |
| AI provider | Claude only | Claude + OpenAI + Gemini + local Ollama |
| Pricing | $20/mo Claude.ai Pro | Solo $468/yr / Professional $948/yr / Firm $1,548/seat/yr + direct API costs (~$5-15/mo typical) |
| Custom instructions per project | Yes | Yes (per workspace + per chat) |
| Project knowledge (file uploads) | Yes | Yes (any folder, any size) |
| Image attachments in chat | Yes, native | Paste, drag, paperclip per provider |
| PDF chat with native vision | Yes | Drop in, Claude reads it natively |
| PDFs as searchable workspace context | Per-project only | Toggle in Memory settings, indexed in LanceDB |
| Long-context [Compress] when you overflow | Hard fail | Auto-trigger or manual button |
| Read aloud (TTS), local | No | Local Piper sidecar |
| Plugin / extension model | No | No plugin ecosystem |
| Searchable across conversations | Within a single project | Full-text across the entire workspace |
| Matter-scoped cited recall | No (per-project only) | Yes (cited answers across all matter docs) |
| Profession-specific workflow packs | No | 50+ built-in templates |
| UI languages | Many (web) | English, Espanol, Deutsch |
| Offline editing | No | Yes (AI calls need network) |
| Multi-device sync | Built-in | BYO via Dropbox / iCloud / Syncthing |
| Mobile | Native apps | Read via cloud sync; native app post-v2.0 |
| Real-time multi-user | No | No |
| Data export | JSON dump | Already in standard files |
| Vendor lock-in | High | Low |
Claude Projects keeps your conversation history on Anthropic's servers, in their cloud, under their retention and data handling policies. That's fine for general-purpose AI work. It's a structural problem for an attorney working under attorney-client privilege, a CPA bound by IRC Section 7216, or a consultant with NDA obligations. The question isn't whether Anthropic is trustworthy. The question is whether your professional obligations permit you to put that client content on any third-party server at all. Advisor Prep Hero answers that question by not being in the data path. Your API key goes straight to Anthropic's inference endpoint. Advisor Prep Hero never sees the request, never logs it, never stores it. The conversation becomes a Markdown file on your machine, where your other client files live, under your own access controls.
Claude Projects only works with Claude. Advisor Prep Hero supports Claude, OpenAI, and Gemini side-by-side, plus local Ollama for the most sensitive content. You pick the model per conversation. If Claude is having an outage, you switch to GPT in one click without losing your workspace.
Advisor Prep Hero ships with 15 practice workflow packs designed specifically for professional work: client intake, engagement letter, matter strategy, weekly review, financial projections, and more. Each runs as a structured interview that produces a real document. Claude Projects has custom instructions but no templated workflows.
Every Advisor Prep Hero conversation is a real Markdown file in a real folder. Back it up with your other client files, include it in a litigation hold, or open it in any text editor five years from now. Claude Projects exports as a JSON dump tied to Anthropic's account. If the account closes or the product changes, the export path is your only lifeline. For professionals with document retention obligations, Advisor Prep Hero's format is already compliant with how you store everything else.
Advisor Prep Hero's matter-scoped recall lets you ask a question and get cited answers drawn from across every document and email in a matter. You can verify each citation against the source. Claude Projects stores conversations per project, but has no equivalent cross-document citation layer; you cannot ask a question and see exactly which source document each piece of an answer came from.
If you draft on your phone during a commute and edit on your laptop later, Claude Projects' native mobile apps and built-in sync are genuinely better. Advisor Prep Hero is desktop-only; mobile sync is BYO via Dropbox or iCloud, which works but isn't real-time.
Claude Projects: sign up at claude.ai, click new project. Advisor Prep Hero: download desktop app, sign up at console.anthropic.com, generate API key, paste it. Both take ~5 minutes; the Claude.ai path is more familiar for someone new to AI tools.
Claude.ai gets new Anthropic features (Artifacts, Computer Use, Skills, etc.) at the moment Anthropic ships them. Advisor Prep Hero supports Anthropic's API capabilities and adds support for new ones over time, but there's a small lag.
| Claude Projects (claude.ai Pro) | Advisor Prep Hero + BYOK | |
|---|---|---|
| Where your conversation history lives | Anthropic's servers, tied to your claude.ai account | Your hard drive. Markdown files in a folder you own. |
| Who handles your AI request | Anthropic (via claude.ai, their consumer product) | Anthropic API directly, via your own API key |
| Does the vendor see your prompts? | Yes, through the claude.ai product layer | Your API key goes straight to Anthropic's inference endpoint. Advisor Prep Hero never sees the request. |
| Default training data policy | Claude.ai consumer product (check current Anthropic terms) | Anthropic API commercial tier: does not train on inputs by default |
| Export format | JSON dump, account-dependent | Already in plain Markdown. Open in any editor. |
| Pricing | $20/mo Claude.ai Pro (flat) | Solo $468/yr / Professional $948/yr / Firm $1,548/seat/yr + your direct API costs ($5-15/mo typical) |
Note: Anthropic's API commercial tier and the claude.ai consumer product have different data handling policies. If confidentiality matters for your use case, read Anthropic's current privacy policy and pay attention to which product tier applies. Advisor Prep Hero operates on the API tier because that's where BYOK lives.
Most professionals don't need to migrate everything. The 5-10 strategically important conversations make sense to bring over; the rest can stay in claude.ai or be discarded.
No. Both work with Claude, but they have different shapes. Claude Projects is a cloud-based workspace inside claude.ai where conversations live on Anthropic's servers. Advisor Prep Hero is a local-first desktop app where every conversation lives on your hard drive as a Markdown file, and you bring your own Claude API key.
No. Advisor Prep Hero uses the Anthropic API tier (BYOK), which is separate from claude.ai. You sign up at console.anthropic.com, generate an API key, paste it into Advisor Prep Hero. You pay Anthropic per token used, typically $5-15/month for professional usage.
Anthropic sees the specific text you send for inference, the same way they would with any API call. They do not see your full conversation history across files; that lives on your machine. Their commercial API does not train on inputs by default.
You can export conversations from claude.ai using the export feature, then convert the JSON to Markdown and drop into your Advisor Prep Hero workspace folder. Advisor Prep Hero reads any folder of Markdown files you point it at.
For API-based features (new models, longer context, tool use), yes. For features that are Claude.ai-specific UI (like the in-product Artifacts panel), Advisor Prep Hero has its own equivalent that ships independently. Both keep up with Anthropic's model releases.
Yes. Many users do. Claude Projects for the chats they want on mobile or shared with a teammate; Advisor Prep Hero for the ones that involve confidential client information they want as a permanent archive on their machine.
Download free for 30 days, no credit card. Solo $468/yr, Professional $948/yr, Firm $1,548/seat/yr (min 3 seats).
Download free for 30 daysFounding offer: 30% off any plan, locked for the life of your subscription. Reserve the founding rate while it lasts.
Built for your field: attorneys, tax pros, consultants.