INDEX / AI_RESEARCH / DEEPWIKI v2026.9
dns archive:data/tools
| | LATENCY: Performance 83/100
Repo wikis Conversational Q&A FREE PRICING Verified Audit

DeepWiki

/ ENGINEERS ONBOARDING ONTO UNFA
business Cognition verified Verified Research Audit

Auto-generated, conversational documentation wikis for GitHub repositories

Composite Index Rating CLASS: A-TIER
85.4/100
Rank #8 in Research & Retrieval
30d Velocity
Benchmark Confidence: v2-5dim rubric
Recorded features: 5 recorded features
Primary Engine: Repo gallery
Citation Precision link
85.0%
Answer grounding against retrieved sources (5-dim quality).
Research Automation hub
76/100
Depth of autonomous multi-source investigation capabilities.
Avg Depth: 5 recorded features
Runtime Performance speed
83/100
Runtime performance & latency dimension (v2-5dim rubric).
Onboarding Usability format_image_left
90.0%
Workflow ergonomics from the v2-5dim rubric.
verified Usability dimension

Recorded Score Vectors

Archive snapshots refreshed on each data update cycle.

rubric v2
menu_book Output Quality & Reliability 85.0%
correctness and consistency of primary task output. v2-5dim
shuffle Feature Depth & Integrations 76.0%
breadth of integrations, APIs, and advanced capabilities. v2-5dim
fact_check Onboarding & Usability 90.0%
onboarding friction, UI clarity, and workflow ergonomics. v2-5dim
attachment Runtime Performance & Latency 83.0%
latency, throughput, and stability under load. v2-5dim
psychology Price-to-Value Efficiency 94.0%
capability delivered per pricing tier dollar. v2-5dim
30-Day Evaluation Quality & Consistency Index
Precision Baseline
Category mean (82/100)
Day -30 (baseline audit) Day -15 (audit refresh) Today (85.4 Composite)

Engine Architecture

Technical primitives powering the DeepWiki pipeline.

01 Repo wikis

Auto-generated wiki-style documentation for any GitHub repository.

Source: DeepWiki · 2026-09-17
02 Conversational Q&A

Ask questions about the codebase and get sourced answers.

Source: DeepWiki · 2026-09-17
03 Repo gallery

Browse popular indexed repos such as vscode, langchain and linux.

Source: DeepWiki · 2026-09-17
04 Devin integration

Index private code with Cognition's Devin.

Source: DeepWiki · 2026-09-17
tune Head-to-Head Comparison

Direct Peer Evaluation Matrix

Updated: Per-audit refresh
Engine / Platform AiRecMark Score Retrieval Paradigm Deep Research Mode Model Agility Hallucination Delta Pricing / Seat
D
DeepWiki Target
● 85.4 Auto-generated, conversational documentation check_circle Agentic research mode Cognition Free for public GitHub repositories
C
Claude
96.1 Adaptive Reasoning Engine check_circle Guided research mode Managed models
P
Perplexity
93.6 Multi-Turn Source Grounding remove Guided research mode Managed models $20/mo Pro
C
ChatGPT
90.9 General Reasoning check_circle Guided research mode Managed models $20/mo
G
Gemini
88.5 Users embedded in Google's ecosystem who wan remove Guided research mode Managed models $19.99/mo
Metrics aggregated from published archive fields. Five-dimension rubric v2 (25/20/15/20/20 weighting).
Financial & Token Econometrics

Subscription & API Tiers

Verified plan specifications and monthly throughput allowances.

Engineering Tier
Pro
Custom / mo
  • check Repo wikis
  • check Conversational Q&A
  • check Repo gallery
  • close Private repo pricing not public (tied to Devin)
  • close Generated docs can lag fast-moving code
Deploy DeepWiki
Recommended
TEAM
Teams
Custom volume
  • check Team seats & workspaces
  • check Volume discounts
  • check Priority support
  • check SSO / SAML & audit controls
Contact Sales
Sovereign Deployment
Enterprise
Custom SLA terms
  • check SSO / SAML & audit controls
  • check Dedicated support channel
  • check Custom quota & SLA
  • close Private repo pricing not public (tied to Devin)
  • close Generated docs can lag fast-moving code
Talk to Sales
verified_user AiRecMark Lead Analyst Assessment
"DeepWiki stands out for public repos browsable free with no paywall. Auto-generated, conversational documentation wikis for GitHub repositories anchors its proposition, and AiRecMark's five-dimension audit lands it at 85.4/100 — a pragmatic default for engineers Onboarding onto Unfamiliar Repos."
EVALUATION SIGN-OFF:
SIGNER: AiRecMark Editorial · 2026-09-17 (AIRECMARK-DAO)
lock_reset
Pricing Source Verification
sha256: 2026-09-17
Export Report PDF