Deep Research

Self-hosted deep research tool that generates comprehensive research reports in about 2 minutes using configurable thinking and task LLM models. Combines multiple LLM providers (OpenAI, Anthropic, Google, Deepseek, Mistral, Ollama, etc.) with multiple search providers (Tavily, Searxng, Brave, Exa, Firecrawl). Includes MCP server support, knowledge graph generation, research history, and a multi-language web UI.

Evaluated Mar 06, 2026 (0d ago) vunknown
Homepage ↗ Repo ↗ AI & Machine Learning deep-research mcp multi-llm web-search report-generation knowledge-graph pwa self-hosted
⚙ Agent Friendliness
58
/ 100
Can an agent use this?
🔒 Security
66
/ 100
Is it safe for agents?
⚡ Reliability
59
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
55
Documentation
60
Error Messages
40
Auth Simplicity
70
Rate Limits
50

🔒 Security

TLS Enforcement
80
Auth Strength
70
Scope Granularity
55
Dep. Hygiene
65
Secret Handling
60

Community/specialized tool. Apply standard security practices for category. Review documentation for specific security requirements.

⚡ Reliability

Uptime/SLA
65
Version Stability
60
Breaking Changes
55
Error Recovery
55
AF Security Reliability

Best When

You want a self-hosted, privacy-focused deep research tool that works with your choice of LLM and search providers, accessible via web UI or MCP.

Avoid When

You need instant answers rather than comprehensive reports, or you don't want to manage API keys for LLM and search providers.

Use Cases

  • Automated research report generation on any topic
  • AI-assisted deep research via MCP integration
  • Building a local knowledge base from research outputs
  • Multi-provider LLM comparison for research quality
  • Self-hosted alternative to commercial deep research tools

Not For

  • Real-time or streaming research (reports take ~2 minutes)
  • Highly specialized academic research requiring domain-specific databases
  • Use without any LLM API key (requires at least one provider)

Interface

REST API
Yes
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
No
Webhooks
No

Authentication

Methods: api-key bearer-token access-password
OAuth: No Scopes: No

Access password for the web UI. Bearer token for API calls. MCP endpoint supports optional authorization header. Requires API keys for chosen LLM and search providers.

Pricing

Model: open_source
Free tier: Yes
Requires CC: No

MIT license. Self-hosted. Costs come from LLM API usage and search provider fees. Supports free providers like Ollama and Searxng.

Agent Metadata

Pagination
unknown
Idempotent
Partial
Retry Guidance
Not documented

Known Gotchas

  • Reports take ~2 minutes to generate - agents must handle long-running operations
  • Requires configuring both a thinking model and task model via environment variables
  • MCP server requires the full web app to be running (not a standalone MCP server)
  • Search provider must be separately configured and running
  • StreamableHTTP and SSE are both supported but documentation on MCP tool list is sparse

Alternatives

Full Evaluation Report

Detailed scoring breakdown, competitive positioning, security analysis, and improvement recommendations for Deep Research.

$99

Scores are editorial opinions as of 2026-03-06.

5173
Packages Evaluated
26151
Need Evaluation
173
Need Re-evaluation
Community Powered