Pinchy vs Glean
Glean is a proprietary enterprise AI platform: search, an assistant and an agent builder across 275+ connectors, run for you by Glean in its own cloud or in your GCP or AWS account. Pinchy is open source (AGPL) and self-hosted: you run it on your own servers, with per-agent permissions and a signed audit trail in the free Community edition. The question is where it runs and who operates it.
The Core Difference
Connect 275+ apps, then search, ask, and build agents that read and write across them. Glean operates the platform as a single-tenant service, either in its own cloud (Glean Hosted) or in your GCP or AWS account (Customer Hosted). Models are cloud-hosted.
Run a single Docker stack on your own servers, with your own models. The center of gravity is governed action: which tools each agent may use, a signed record of every call, and a core that is free at any size.
Glean is closed-source and operated by Glean, including in Customer Hosted mode. Pinchy is AGPL-3.0: you can read the code, change it, patch it on your schedule, and run it air-gapped. No vendor sits between you and the software.
Side by Side
Figures as of 16 October 2026, taken from Glean's public pricing, product and documentation pages. Check Glean for current details.
| Glean | Pinchy | |
|---|---|---|
| Category | Enterprise AI platform: search, assistant, agents | Self-hosted agent platform |
| License | Proprietary (closed source) | AGPL-3.0 (open source) |
| Who operates it | Glean (single-tenant, in its cloud or in your GCP/AWS) | You (Docker, your servers) |
| Inspect / modify the code | No | Yes |
| Free tier | No (sales-led, enterprise) | Yes (AGPL, free) |
| Pricing | Custom pricing (not published). Per user per month, plus usage (Model Hub commit or FlexCredits) | Community free (AGPL), no user limit; Pro €1,188/yr for up to 10 users, quote above |
| Where it runs | Glean Hosted, or Customer Hosted in your GCP/AWS | Your own servers, one Docker Compose stack |
| Local models / offline | Cloud-hosted LLMs (your own keys possible); no local or offline option | Any provider via your keys, or Ollama fully offline |
| Connectors | 275+ connectors with read & write tools, plus MCP | Odoo (read + write), email (Gmail, Microsoft 365, IMAP), web search, knowledge base |
| Per-agent tool permissions | Role-based access controls for agents (Agent Governance, included in the seat) | Default-deny allow-list per agent: no tools until an admin grants them |
| Audit of agent actions | Audit trail of control points, streamable to a SIEM | Every tool call and knowledge-base retrieval; each row HMAC-SHA256-signed and hash-chained, built-in verifier, CSV export, free |
Being Honest
Glean runs a hybrid search index and a knowledge graph over 275+ connectors, with permissions enforced per connector. For company-wide search across everything, it operates at a scale Pinchy does not target. Pinchy's knowledge base covers folders you grant per agent, indexed locally, with answers that cite their source page.
Slack, Teams, Zoom, Zendesk, GitHub, Confluence, Salesforce and 250+ more, with read and write tools and MCP support. Pinchy reaches Odoo, email, web search and your knowledge base, and is deliberately focused. Slack and Teams are planned, not available.
Glean lists SOC 2 Type II, ISO 27001, ISO 42001, HIPAA and GDPR on its site, offers single-tenant deployment, and streams its audit events to a SIEM. For enterprise procurement, that track record matters. Pinchy holds no such certifications.
Glean runs the platform for you, even in your own cloud. If you do not want to operate infrastructure and have the budget, that is a real advantage over self-hosting Pinchy.
Where Pinchy Wins
Pinchy is AGPL-3.0. Read the code, modify it, fork it, patch on your own schedule. With Glean, you depend on a proprietary vendor for every change, even when it runs in your cloud.
One Docker Compose stack on your own servers, your own models. With a local model via Ollama, prompts and documents stay on your network; only the external integrations you grant (web search, Gmail or Microsoft 365, Telegram) reach outside. With a local model and no external integrations, it runs fully air-gapped. No telemetry, no phone-home, no vendor operating your deployment. In return you run the server and pick the model.
The allow-list per agent, the signed audit trail, the integrations and offline operation are in the free Community edition, with no user limit. Pro adds groups, agent visibility and usage analytics: €1,188/yr for up to 10 users, quote above. Pinchy's licence covers neither your servers nor model usage.
Agents get no tools beyond their own workspace until an admin grants them, per agent; templates propose a starting set, and only admins create agents. Every tool call is recorded in an audit row that is HMAC-SHA256-signed and hash-chained. A built-in verifier checks the chain and you can export it as CSV — in the free tier. Tamper-evident, not tamper-proof.
Pinchy ships a deep Odoo integration: query stock, create orders, reconcile invoices, all from chat or Telegram. Acting inside an ERP is a first-class use case, not one connector among hundreds.
Pinchy does not pause to ask. The control is the grant. Give an agent email_draft but not email_send and it can prepare a mail but never send it. Give it Odoo read tools without write tools and it can look but not change. An admin decides, per agent, what it may do.
Decision Guide
You need company-wide AI search, assistants and agents across 275+ SaaS apps, want a vendor-operated single-tenant deployment, and have an enterprise budget.
You want open-source agents you run on your own servers, with per-agent governance and a signed audit, reaching Odoo, email and your documents — and you accept running a server and choosing a model.
Do you want a vendor-operated platform that reaches 275+ apps, with models in the cloud? That is Glean. Or governed agents on servers you own, offline if you choose? That is Pinchy.
FAQ
For teams that want open-source agents they run on their own servers, yes — though the two differ in scope. Glean is a proprietary enterprise AI platform (search, assistant and agents across 275+ connectors), operated by Glean as a single-tenant service. Pinchy is an AGPL, self-hosted agent platform you run on your own infrastructure, with per-agent permissions and a signed, hash-chained audit trail in its free Community edition.
No. Glean is a proprietary, closed-source platform. It runs as a single-tenant service, either in Glean's cloud (Glean Hosted) or in your GCP or AWS account (Customer Hosted), but Glean operates the deployment and you cannot inspect or modify the code. Pinchy is AGPL-3.0: you run it yourself with a single Docker Compose stack and can audit and change every part of it.
Glean does not publish prices: custom pricing. Its documentation describes licensing per user per month, with model usage billed separately as a Model Hub commit or as FlexCredits. Pinchy Community is free under AGPL-3.0 with no user limit. Pro adds groups, agent visibility and usage analytics for €1,188 per year (excl. VAT) up to 10 users, with a quote above that. Pinchy's licence covers neither your servers nor model usage.
Glean runs single-tenant, in Glean's cloud or in your GCP or AWS account, and Glean operates the deployment either way. Its models are cloud-hosted. Pinchy runs on infrastructure you operate yourself. With a local model via Ollama, prompts and documents stay on your network, and only the external integrations you grant reach outside. The license is validated offline, so Pinchy never phones home.
Choose Glean when you need company-wide AI search, assistants and agents across 275+ apps, run for you as a single-tenant service, and you have an enterprise budget. Choose Pinchy when you want open-source agents on servers you operate, with per-agent governance and a signed audit, reaching Odoo, email and your documents, and you accept running a server and picking a model.
Self-host Pinchy yourself in minutes, or book a call to talk it through. Your choice.
Or email us: info@heypinchy.com