Resources / Coverage index

Hundreds of AI providers. One reviewed detection catalog.

Tallin Discover identifies supported AI services from managed-browser hostname evidence. This index shows what the catalog covers, how a provider enters it, and where hostname discovery stops.

Current release
123 reviewed provider identities
Hostname claims
133 evidence-backed match claims
Verification window
Aug 10, 2026 to Aug 10, 2026
Candidate intake
Daily scan plus customer requests and Tallin research
Publication
Human review, code review, safety gates, versioned release

Counts are generated from the production catalog at build time. They update when a reviewed catalog release ships, so this page does not rely on a manually maintained marketing number.

Scope

The catalog makes one bounded claim.

Each catalog entry states that a reviewed hostname belongs to a named AI provider. A matching browser event establishes that the supported service was reached from a managed browser context. It does not reveal the prompt, response, URL path, search, page content, purpose, or provider account.

Whether that use is sanctioned depends on the customer’s approved provider list, enterprise workspace roster, identity evidence, and policy. The same service can be sanctioned for one organization and unresolved for another. Tallin keeps the provider identity separate from the customer’s governance decision.

Current provider coverage

Browse the production catalog by category.

Provider names are public. Raw matching patterns stay in the reviewed catalog so changes continue through the same safety and release process as product code.

Agent platforms8 providers
  • Botpress
  • Browse AI
  • Dust
  • Gumloop
  • Kore.ai
  • Manus
  • Relevance AI
  • Voiceflow
AI assistants4 providers
  • Genspark
  • MaxAI
  • Merlin
  • Monica
AI gateways1 provider
  • Portkey
AI search2 providers
  • Perplexity
  • You.com
Chat and general assistants13 providers
  • Abacus.AI / ChatLLM
  • Anthropic / Claude
  • Character.AI
  • Cohere
  • DeepSeek
  • Duck.ai
  • Grok
  • Kimi (Moonshot)
  • Meta AI
  • Mistral AI
  • OpenAI / ChatGPT
  • Pi
  • Poe
Coding assistants14 providers
  • Augment Code
  • BLACKBOX AI
  • Bolt
  • Continue
  • Cursor
  • Devin
  • Factory
  • Lovable
  • Phind
  • Pieces
  • Qodo
  • Tabnine
  • v0
  • Windsurf / Codeium
Customer service AI3 providers
  • Decagon
  • Forethought
  • Sierra
Data analysis1 provider
  • Julius AI
Document AI3 providers
  • ChatPDF
  • Humata
  • PDF.ai
Enterprise assistants3 providers
  • Aisera
  • Moveworks
  • Sana Agents
Enterprise research1 provider
  • Hebbia
Enterprise search1 provider
  • Glean
Government AI1 provider
  • Ask Sage
Image generation8 providers
  • Adobe Firefly
  • Ideogram
  • Krea
  • Leonardo.Ai
  • Midjourney
  • Playground
  • Recraft
  • Stability AI
Legal AI2 providers
  • Harvey
  • Spellbook
Meeting assistants12 providers
  • Avoma
  • Circleback
  • Fathom
  • Fireflies.ai
  • Grain
  • Granola
  • Krisp
  • MeetGeek
  • Otter.ai
  • Read AI
  • Tactiq
  • tl;dv
Model and ML platforms9 providers
  • Baseten
  • Cerebras Cloud
  • fal
  • Fireworks AI
  • Groq
  • Hugging Face
  • OpenRouter
  • Replicate
  • Together AI
Music generation2 providers
  • Suno
  • Udio
Presentation AI3 providers
  • Beautiful.ai
  • Gamma
  • Napkin AI
Productivity AI4 providers
  • Google Gemini / Bard
  • Google NotebookLM
  • Mem
  • Microsoft Copilot (Bing)
Research AI4 providers
  • Consensus
  • Elicit
  • SciSpace
  • scite
Video generation10 providers
  • Descript
  • Hailuo AI
  • HeyGen
  • InVideo AI
  • Kling AI
  • Luma AI
  • OpusClip
  • Pika
  • Runway
  • Synthesia
Voice agents2 providers
  • Retell AI
  • Vapi
Voice generation3 providers
  • ElevenLabs
  • Murf AI
  • Resemble AI
Writing assistants9 providers
  • Copy.ai
  • Grammarly
  • Jasper
  • Jenni AI
  • QuillBot
  • Sudowrite
  • Wordtune
  • Writer
  • Writesonic

Methodology

Candidate first. Evidence next. Publication last.

01

Candidate intake

AI-categorized Defender, Netskope, and approved network or CASB evidence can propose unknown hostnames. Customer requests and Tallin research enter the same queue.

02

Vendor evidence

A reviewer verifies the exact hostname against a vendor-controlled HTTPS source and records when the claim was added and last checked.

03

Safety gates

Build tests reject duplicate, overlapping, malformed, lookalike, and overbroad claims before they can affect any customer.

04

Versioned release

A human-approved code change publishes the provider. The browser receives a checksum-verified, content-versioned catalog bundle.

Known limits

Coverage is named, not implied.

Embedded AI
An AI feature inside a general-purpose SaaS hostname may be indistinguishable from ordinary use of that product. Tallin does not classify the whole application as AI to inflate coverage.
Native clients
Desktop and mobile AI applications can bypass managed-browser navigation. Network, provider, or endpoint evidence may be required.
Unmanaged use
Personal devices, unmanaged browsers, and off-network paths remain outside browser coverage unless another configured source observes them.
Provider account
A hostname observation does not prove whether the employee used a personal, free, or approved enterprise account.
Unknown services
A provider outside the reviewed catalog is not reported by the privacy-limited browser channel. Candidate sources can surface it for research without collecting general browser history.

References

  1. [1]
    NIST Generative AI Profile

    National Institute of Standards and Technology

  2. [2]
    chrome.webNavigation API

    Chrome for Developers

  3. [3]

See which AI services appear across your managed browsers.

Start Tallin Discover with privacy-limited hostname evidence, sanctioned-access reconciliation, and a review queue built for owned decisions.

Shadow AI Provider Coverage Index | Tallin