Cherry is the self-hosted alternative to Poe.
Poe is Quora's hosted aggregator giving consumers access to GPT, Claude, Gemini, image, video and audio models plus a large catalog of user-created bots. It's free at low usage, with subscriptions starting at $4.99/mo for advanced usage. It's a great catalog and a fine consumer surface. It's also SaaS on someone else's platform, with no folder tree, no per-prompt cost attribution and no workspace roles. Cherry is the professional alternative: a fifteen-plus entity Postgres data model, the Pins/Clipboard/Snippets three-tier reuse system, per-message UsageEvent attribution across user / workspace / chat / model, private MinIO file storage with presigned URLs, real multi-workspace teams with roles and modular Apps. Self-hosted end to end. You own the whole stack.
Four things a bot catalog can't do for a working team.
Poe solves a real problem for casual users: one login, dozens of models, thousands of user-created bots. It doesn't solve the workspace problem sitting underneath a professional's AI day.
Consumer bot aggregator
Poe optimizes for browsing a large catalog of user-created bots. That's great for discovery and useless for organizing a real workstream. Cherry ships a self-referential Folder tree with unbounded nesting, Pins on individual messages, a per-chat Clipboard and project-wide Snippets with tags. The workspace treats your work as work, not as bot exploration.
Quora owns the platform
Poe is Quora's product. Every chat, every bot conversation, every file lives on Quora's infrastructure. If Poe pivots, sunsets a model, changes credit economics or moves against a category, your work goes with it. Cherry is self-hosted end to end. Postgres, Redis and MinIO run on your hardware. There's no third-party pivot risk.
Subscription hides real cost
Poe subscriptions bundle model access into monthly tiers with compute-point limits behind the scenes. You never know what a single research thread actually cost. Cherry logs a UsageEvent per message with tokens, dollar cost and latency attributed to user, workspace, chat and model. A consultant attributes AI spend per client engagement in one query.
No workspace roles
Poe is designed around an individual account. Group chat exists but there's no workspace hierarchy, no Owner/Admin/Member roles, no per-workspace billing plans, no storage attribution per team. Cherry ships multi-workspace with roles, per-workspace BillingPlan and StorageUsage entities.
A workspace built for work, not a bot store built for discovery.
Poe optimizes for browsing a huge catalog of user-generated bots on Quora's platform. Cherry optimizes for the day after you've picked your models. A curated Agent Library replaces community bot discovery. A nested Folder tree replaces a flat chat list. UsageEvents replace opaque compute-point subscriptions. MinIO on your infrastructure replaces Quora's file storage. Multi-workspace with roles replaces per-account access. And the whole thing runs on your hardware, not on someone else's product strategy.
Poe vs Cherry.
A consumer aggregator with a bot store versus a professional workspace with a data model.
| Feature | Poe | Cherry |
|---|---|---|
| Owned by | Quora | You |
| Deployment | SaaS at poe.com | Self-hosted. On-prem, cloud, run as your own SaaS. You decide |
| Model catalog | GPT, Claude, Gemini, Grok, DeepSeek, image, video, audio, large user-bot catalog | OpenAI + Anthropic shipping, typed adapter for any provider |
| Pricing | Free + subscriptions from $4.99/mo (compute points) | Contact sales, self-hosted licensing |
| Model presets | Bots (community + custom) | Agent Library with model + temp + prompt + tier gating |
| Organization | Chat list | Infinite nested folder tree, self-referential parents |
| Per-message bookmarks | No | Pins with scroll-to-highlight animation |
| Per-chat clipboard | No | Drag-reorderable saved passages, promote to Snippet |
| Cross-chat snippets | No first-class equivalent | Tagged, folder-scoped, one-click injection to any chat |
| Model comparison | Multi-bot chat | Compare modal fires one prompt at every provider, diff answers |
| Cost per prompt | Compute-point consumption | Tokens, dollar cost, latency on every message |
| Cost attribution | Account-level subscription | Per user, per workspace, per chat, per model |
| Usage dashboard | Point balance | Recharts. Today / 7d / 30d / billing cycle |
| File storage | Quora infrastructure | Self-hosted MinIO with local-disk fallback, presigned URLs |
| Bulk folder upload | No | Replicates an entire OS folder tree transactionally |
| Conversation history | Quora-owned | Your Postgres. Exportable, portable |
| Streaming transport | Proprietary | Server-Sent Events with cancel-mid-stream |
| Team workspaces | Not first-class | Multi-workspace with Owner / Admin / Member roles |
| SSO | Consumer login | Available in Enterprise, configurable on your infrastructure |
| Outbound API | Bot creator API | Cherry outbound API for third-party integration |
| Ownership | Vendor-owned | You own the code, the data, the deployment |
Frequently asked questions.
Can Cherry match Poe's bot catalog?
Cherry doesn't try to. Poe's value is discovery across a large catalog of user-generated bots. Cherry's Agent Library is a curated Preset system with model + temp + system prompt + tier requirement, meant for a professional team to build and maintain its own working set. Different problem, different design.
Which models does Cherry ship?
OpenAI and Anthropic adapters ship today. The provider abstraction is a typed interface, so wiring Google, xAI, DeepSeek, Perplexity Sonar or any future provider is a single-file adapter addition. Google models are already in the catalog for Preset filtering; the runtime adapter is on the roadmap.
Does Cherry do image, video and audio like Poe?
Cherry ships an Image Generation App and a Canvas app for live-rendering generated code. Video and audio generation aren't first-class today. The provider adapter interface accepts any future modality, so wiring Sora, Kling or ElevenLabs is a config addition, not a rearchitecture.
How granular is the cost tracking?
Every message generates a UsageEvent record with tokens in, tokens out, dollar cost, latency in ms, model, preset, chat, user and workspace. The Recharts dashboard filters by today, 7 days, 30 days or billing cycle and shows a daily cost bar chart, breakdown by user / model / chat and storage growth in GB.
How is Cherry deployed?
Cherry is self-hosted. Docker Compose orchestrates nginx, the Next.js frontend, the Express backend, Postgres, Redis and MinIO. Deploy it on your own hardware for internal use, run it in your cloud or spin it up as your own SaaS product. You own the code, the data and the deployment.
What does Cherry cost?
Contact sales for pricing. Every engagement is scoped to how you plan to deploy, who your users are, which provider adapters you need lit up and what support terms fit your organization.
Ready to work in a real workspace instead of a bot catalog?
Tell us how your team uses Poe today, which models and bots you rely on and where the consumer surface starts costing you. We'll respond within one business day with an honest read on whether Cherry fits.