Odysseus AI
odysseusai.org
Odysseus AI — Private AI Research & Agent Workspace
Description
Odysseus AI: The Private AI Workspace for Agent Workflows and Deep Research. · A private AI workspace for deep research and autonomous agents. Turn files into source-traceable reports with OpenRouter BYOK and custom MCP tools. · Automated intelligence for high-stakes analysis. · From private context to finished AI work. · One isolated workspace for serious AI workflows. · See deep research turn your sources into cited reports.
Odysseus AI is a private AI workspace built for teams and builders who need control. Today, it turns PDFs, URLs, and notes into source-traceable cited research briefs via OpenRouter routing with optional BYOK. Autonomous agents with MCP, memory, and email/calendar operations are on the roadmap.
Features
Private AI Research Workspace
PDF & URL Ingestion to Cited Research Briefs
Source-Traceable Research Reports
OpenRouter Routing with BYOK (Bring Your Own Key)
Isolated Workspace — No Model Training on User Data
Research Agent Workflow with MCP Tools
Use cases
Academic Literature Review
Competitor Benchmarking & Market Intelligence
Due Diligence & Investment Research
Private Research Workflows for Teams
Multi-Source Research Automation
FAQ
Odysseus AI is a private AI research workspace that allows users to upload PDFs, URLs, and notes to generate structured, source-traceable research briefs. It's designed for professionals who need control over their data and a verifiable audit trail for their AI-assisted work.
Odysseus AI provides isolated cloud workspaces and does not train its models on user data. It offers optional Bring-Your-Own-Key (BYOK) via OpenRouter for model access and is evaluating self-hosted deployment paths. Data retention is minimized, and no behavioral telemetry is collected by default.
It's ideal for academic literature reviews, competitor and market analysis, investment due diligence, and product benchmarking. Any workflow requiring structured research output from multiple source documents while maintaining data privacy is a strong use case.
Not in the current Preview. The platform uses managed routing through OpenRouter. Self-hosted and local model execution paths are explicitly stated as being under evaluation for future releases.
Plans differ primarily in the number of monthly research reports (20, 60, 150), saved source/note capacity (500, 2.5k, 10k), and workspace storage (25GB, 100GB, 250GB). All include managed model usage under fair use policies and privacy features.
No. The current Preview focuses on guided research workflows where the user controls the context. Agent workflows with controlled access to files and tools are a feature on the product roadmap, not available in the initial release.
Access is currently managed through an early access waitlist. You can request early access via the 'Request Early Access' button or link on the website's pricing or homepage sections.
The website links to a GitHub repository with significant stars and forks, suggesting an open-source component or related project. However, the main managed-cloud workspace product described appears to be a commercial service.
The pricing page does not specify overage charges. It lists fixed quantities of research reports per month for each tier, implying that users are limited to those amounts within their billing cycle.
Odysseus AI emphasizes source traceability with a detailed citation table, structured report output, and a stronger focus on data privacy with BYOK and isolated workspaces. It is also built with a roadmap towards integrated agents, email/calendar, and MCP tools, positioning it as a broader private AI workspace.
Specs
| Type | Agent |
| Section | Productivity |
| Pricing | paid |
| Platform | API only |
| Systems | api, web |
| Who for | команды и бизнес |
| Site language | en |
| Rating | 0.00 (0 reviews) |
| Views | 63 |
| Launched | 2026-06-25 |
