Pinchy vs Google Agentspace
Google Agentspace, which Google now offers as part of its Gemini Enterprise lineup, is a cloud-hosted enterprise agent and search product built around Gemini and the Google Workspace ecosystem. Pinchy is self-hosted first: AGPL open source, model-agnostic, with default-deny per-agent permissions and a signed audit trail in the free tier. Pick the one that matches where your data must live.
The Core Difference
Agentspace connects your Google data and SaaS sources, runs enterprise search across them, and builds agents on Gemini. The center of gravity is Google Cloud: hosted, operated, and billed by Google, with deep Workspace integration.
Pinchy runs on your own infrastructure: one Docker Compose stack, your models, your data. The center of gravity is governance and sovereignty: which tools each agent may use, and a signed record of everything it does, with nothing leaving your network.
Agentspace is proprietary and built around Gemini. Pinchy is AGPL-3.0 open source and model-agnostic: Anthropic, OpenAI, Google, or local Ollama. Neither hosting nor model is locked in at the platform level.
Side by Side
Figures as of June 2026. Check Google Cloud and Pinchy docs for current details.
| Google Agentspace | Pinchy | |
|---|---|---|
| Product model | Cloud service on Google Cloud | Self-hosted first (AGPL) |
| Source | Proprietary | Open source (AGPL-3.0) |
| Free tier | No (paid editions) | Yes (AGPL, no user limit) |
| Pricing model | Per-user editions plus consumption-based usage | Free plus flat Pro (not per seat) |
| Where data is processed | Google Cloud | Your infrastructure |
| Self-hosting / air-gapped | No | First-class (Docker Compose, GHCR) |
| Models | Gemini-centric | Anthropic, OpenAI, Google, Ollama (local) |
| Enterprise search breadth | Deep, across Google and SaaS sources | Document knowledge base, narrower |
| Workspace integration | Native Gmail, Drive, Docs, Calendar | Gmail/email, Odoo, web, docs |
| Channels | Web, Workspace surfaces | Web UI plus Telegram |
| Per-agent tool permissions | Google Cloud IAM and policies | Default-deny allow-list per agent |
| Audit of agent actions | Google Cloud audit logging | HMAC-signed per row, free, CSV export |
| SSO / SAML, granular RBAC | Yes (Google Cloud IAM, Workspace) | On the roadmap, not yet shipped |
| Scale / backing | Google Cloud | Early-stage, built in public |
Being Honest
Agentspace runs unified enterprise search across Google data and connected sources like Drive, Jira, Confluence, SharePoint, and Salesforce, synthesizing answers with links back to the source. Pinchy's document knowledge base is deliberately narrower.
If your company lives in Gmail, Drive, Docs, and Calendar, Agentspace has deep, native access to that data and can act on it inside the Google ecosystem. Pinchy integrates Gmail and email but is not built around Workspace.
Agentspace is tuned for Gemini and runs on Google-scale infrastructure, with the models, throughput, and support that implies. If Gemini is your model of choice, that tight coupling is a strength rather than a constraint.
Agentspace inherits Google Cloud IAM, SSO and SAML, granular role-based access, and Google's compliance posture out of the box. Pinchy's SSO and granular custom RBAC are on the roadmap, not yet shipped.
Where Pinchy Wins
Pinchy runs entirely on your infrastructure. Paired with a local model via Ollama, no prompt or document ever leaves your network, air-gapped if you need it. There is zero telemetry and no phone-home. With Agentspace, data is processed in Google Cloud.
Choose Anthropic, OpenAI, Google, or a local Ollama model per agent, and switch without re-platforming. Agentspace is built around Gemini. If your model strategy must stay open, that difference matters.
Pinchy is free under AGPL with a flat Pro tier (around EUR 99 per month, up to 10 users, not per seat). Agentspace bills per user per month across editions, plus consumption-based costs for the underlying agent and model usage, so a team's bill scales with both headcount and usage.
Each agent starts with zero tools; admins enable each one explicitly, per agent. Every tool call is recorded in an individually HMAC-SHA256-signed audit row you can export to CSV and verify. No shell, no code execution, no unattended runs.
Pinchy is AGPL-3.0 and public on GitHub. Security teams can read the whole stack, fork it, and run it without trusting a vendor's black box. Agentspace is proprietary by design.
Pinchy ships a deep Odoo integration over read-write JSON-RPC: query stock, create orders, reconcile invoices, all from chat or Telegram. Acting inside an ERP is a first-class use case, not a generic connector.
Decision Guide
You are already all-in on Google Cloud and Workspace, you want enterprise search across Google and SaaS data, Gemini is your model of choice, and processing data in Google Cloud under per-user Google Cloud billing is acceptable.
Data must stay on your own infrastructure or fully offline, you want to choose your own models instead of Gemini, you want default-deny per-agent governance and a signed audit without per-user cloud billing, and especially when Odoo is central to the work.
Where must the data live, and which models must you be able to run? If the answer is "our servers" and "any provider, including local", that points to Pinchy. If it is "Google Cloud is fine" and "Gemini plus Workspace search", that points to Agentspace.
FAQ
No. Google Agentspace, which Google now offers as part of its Gemini Enterprise lineup, is a proprietary Google Cloud service. You configure and use it, but you cannot read or modify the source. Pinchy is open source under AGPL-3.0 and public on GitHub, so you can audit, fork, and run the whole stack yourself.
No. Agentspace is a cloud product that runs on Google Cloud infrastructure, and your prompts and connected data are processed there. There is no self-hosted or air-gapped deployment. Pinchy is self-hosted first: a single Docker Compose stack on your own servers, and paired with a local model via Ollama no request ever leaves your network.
For teams that need data to stay on their own infrastructure, want to choose their own models instead of being tied to Gemini, and want governance and a signed audit in the free tier, yes. Agentspace is the better fit if you are already all-in on Google Cloud and Workspace and want enterprise search across that ecosystem. Pinchy is the sovereign, model-agnostic, self-hosted path.
Agentspace, now offered through Google's Gemini Enterprise editions, is a paid Google Cloud product. Pricing is published per user per month across tiered editions, on top of which the underlying agent and model usage adds consumption-based Google Cloud costs. There is no free tier. Pinchy is free under AGPL-3.0 with no user limit, and adds a flat Pro tier (around EUR 99 per month billed annually, up to 10 users, not per seat). Roles and the signed audit trail are in the free tier. Check Google Cloud pricing for current figures.
Teams with data-residency or sovereignty requirements that cannot send prompts and documents to Google Cloud, teams that want to run local or non-Google models (Anthropic, OpenAI, Ollama), and teams that want default-deny per-agent permissions plus a tamper-evident audit trail without per-user cloud billing. If Google Workspace is the center of your work and Gemini is your model of choice, staying on Agentspace is reasonable.
Self-host Pinchy yourself in minutes, or book a call to talk it through. Your choice.
Or email us: info@heypinchy.com