Category evidence snapshot

AI Infrastructure Trust Evidence Snapshot 2026

A public evidence snapshot of trust, security, privacy, compliance, and enterprise-readiness surfaces across selected standalone AI infrastructure vendors and serious project entities.

Public evidence only. Signals are interpretive and do not by themselves prove buying intent, non-compliance, procurement pressure, or active vendor search.

856 candidate entities triaged
150 priority evidence scans
32 public standalone examples

What this category includes

This snapshot covers AI infrastructure and LLMOps-related vendors such as model hosting, inference, vector database and retrieval, LLM gateways, AI observability, evaluation, developer infrastructure, and open-source-commercial infrastructure projects.

Evidence surfaces tracked

  • Trust Centers, security pages, and customer assurance surfaces.
  • Privacy, legal, terms, DPA, and subprocessor pages.
  • SOC 2, ISO, HIPAA, GDPR, and enterprise-readiness statements when publicly available.
  • AI data usage, model training, no-training, docs, status, and admin-control evidence.

Exclusion rule

Product aliases, model names, parent-vendor product labels, weak launch labels, and unresolved entities were excluded from named public snapshot examples. They remain internal triage or parent-mapping records.

Snapshot inclusion requires a standalone vendor or serious project entity, verified official website, and at least one official evidence surface.

Representative public evidence patterns

The public sample intentionally includes mature, partial, and sparse evidence patterns. It is not a ranking and does not claim that any vendor is more secure, more compliant, or currently buying trust tooling.

18 benchmark examples
9 partial / evolving examples
5 sparse / gap examples

Benchmark evidence examples

Standalone AI infrastructure vendors with multiple mature public trust, security, legal, privacy, compliance, or enterprise-readiness evidence surfaces.

LangChain

AI evaluation / observability

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal evidence surface for AI infrastructure review.
Observed surfaces
Homepage; Trust Center; Privacy Policy; Terms of Service; status page; enterprise agent platform positioning observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Baseten

model deployment / fine-tuning

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal evidence surface for AI infrastructure review.
Observed surfaces
Homepage; Trust Center; Privacy Policy; Terms with DPA annex; SLA; certification or compliance-related public language; careers page observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

OpenAI

foundation model / model provider

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal evidence surface for AI infrastructure review.
Observed surfaces
Trust Portal; Security & Privacy page; Privacy Policy; Services Agreement; DPA; enterprise privacy/data controls; certification or compliance-related public language; admin controls observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Pinecone

AI database / vector data layer

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal evidence surface for AI infrastructure review.
Observed surfaces
Homepage; Security page; security center link; Privacy; Terms; certification or compliance-related public language; SSO/RBAC/CMEK/private endpoints observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Langfuse

AI evaluation / observability

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal evidence surface for AI infrastructure review.
Observed surfaces
Security & Compliance overview; certification or compliance-related public language; DPA; subprocessors; data regions; RBAC/SSO/SCIM/audit logs; privacy; terms observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Weights & Biases

AI evaluation / observability

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal evidence surface for AI infrastructure review.
Observed surfaces
Homepage; security page; privacy; terms; compliance-ready enterprise statements; certification or compliance-related public language; deployment controls observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

n8n

AI infrastructure / platform

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal evidence surface for AI infrastructure review.
Observed surfaces
Homepage; Legal hub; Trust Center; Security; privacy; DPA; AI terms; certification or compliance-related public language; enterprise automation positioning observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Bland AI

AI infrastructure / platform

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal evidence surface for AI infrastructure review.
Observed surfaces
Homepage; Trust & security portal; Terms; Privacy; DPA; Subprocessors; certification or compliance-related public language; data residency; access controls; regulated-industry positioning observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Comet

AI evaluation / observability

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal evidence surface for AI infrastructure review.
Observed surfaces
Homepage; Opik/AI observability; Security & Compliance link; Terms; Privacy; enterprise-grade reliability/security claims; careers observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Groq

AI infrastructure / platform

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal evidence surface for AI infrastructure review.
Observed surfaces
Homepage; GroqCloud; Trust Center; Security; Privacy Policy; website terms; GroqCloud terms; Services Agreement; enterprise inquiry; careers observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Weaviate

AI database / vector data layer

benchmark evidence mature
Evidence pattern
Mature public trust/security evidence surface for vector database and retrieval infrastructure review.
Observed surfaces
Homepage; security page; Trust Portal; public Trust Portal launch post; enterprise cloud and deployment controls; privacy page; certification-related public language requiring verification observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Modal

AI cloud compute / model execution

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal evidence surface for AI cloud compute and model execution review.
Observed surfaces
Homepage; Trust Center; security docs; privacy policy; DPA; subprocessor update feed; certification or compliance-related public language; workload isolation and enterprise security references observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Fireworks AI

model serving / inference infrastructure

benchmark evidence mature
Evidence pattern
Mature public trust/security and AI data-handling evidence surface for inference infrastructure review.
Observed surfaces
Homepage; Trust Center; security and compliance docs; privacy policy; data security docs; data handling docs; certification-related public language; zero data retention and enterprise inference positioning observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Scale AI

AI data infrastructure / model services

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal evidence surface for AI data infrastructure and model-services review.
Observed surfaces
Homepage; Trust Center; security page; privacy policy; subprocessor page; product terms; certification or compliance-related public language; enterprise and government AI positioning observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Anthropic

foundation model / model provider

benchmark evidence mature
Evidence pattern
Mature public trust/security and AI governance evidence surface for foundation model provider review.
Observed surfaces
Homepage; Trust Center; privacy policy; DPA and subprocessors via Trust Center resources; certification-related public language; Frontier AI compliance and model documentation; enterprise and government offerings observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Cohere

foundation model / model provider

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal and enterprise data-commitment evidence surface for foundation model provider review.
Observed surfaces
Homepage; Trust Center; security page; privacy policy; legal center; enterprise data commitments; certification or compliance-related public language observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Supabase

AI data infrastructure / developer data platform

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal evidence surface for developer data infrastructure and AI application backend review.
Observed surfaces
Homepage; Trust Center; security page; certification or compliance-related public language; privacy policy; DPA; subprocessor access via Trust Center; certification-related public language; backend/database controls observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Vercel

AI app deployment / developer cloud platform

benchmark evidence mature
Evidence pattern
Mature public trust/security/legal evidence surface for AI application deployment and developer cloud review.
Observed surfaces
Homepage; Security portal; security page; compliance docs; privacy notice; DPA; v0 enterprise addendum; subprocessor references; certification-related public language; AI-product legal/security change candidates observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Partial / evolving trust surface examples

Standalone vendors with official trust evidence, but where the public surface is distributed, incomplete, or still evolving.

Hugging Face

AI infrastructure / platform

partial public trust surface
Evidence pattern
Official evidence present, but trust surface is distributed, incomplete, or still evolving.
Observed surfaces
Homepage; privacy; terms; Hub security docs; enterprise plan with SSO, regions, audit logs, resource groups; status observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Replicate

model deployment / fine-tuning

partial public trust surface
Evidence pattern
Official evidence present, but trust surface is distributed, incomplete, or still evolving.
Observed surfaces
Homepage; enterprise page; privacy; terms; enterprise-grade security/compliance claims; DPA availability stated on enterprise page; model API/fine-tuning/custom deployment surfaces observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Mistral AI

foundation model / model provider

partial public trust surface
Evidence pattern
Official evidence present, but trust surface is distributed, incomplete, or still evolving.
Observed surfaces
Homepage; legal center; privacy policy; commercial terms; DPA; trust center referenced by DPA; enterprise/self-hosted/EU cloud/data privacy positioning; compute/inference infrastructure observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Not Diamond

model deployment / fine-tuning

partial public trust surface
Evidence pattern
Official evidence present, but trust surface is distributed, incomplete, or still evolving.
Observed surfaces
Homepage; privacy; legal page; certification or compliance-related public language; ZDR policy statement observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Ollama

LLM developer tooling

partial public trust surface
Evidence pattern
Official evidence present, but trust surface is distributed, incomplete, or still evolving.
Observed surfaces
Homepage; privacy policy; terms; local/offline deployment positioning; no-training and transient cloud processing statements; legal pages current to 2026 observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Activeloop

AI database / vector data layer

partial public trust surface
Evidence pattern
Official evidence present, but trust surface is distributed, incomplete, or still evolving.
Observed surfaces
Homepage; docs; BYOC/BYOM/private deployment/security positioning observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Firecrawl

LLM developer tooling

partial public trust surface
Evidence pattern
Official evidence present, but trust surface is distributed, incomplete, or still evolving.
Observed surfaces
Homepage; privacy policy; terms; security link; certification or compliance-related public language; status page; web data/API infrastructure for agents and LLM applications observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Shaped

AI database / vector data layer

partial public trust surface
Evidence pattern
Official evidence present, but trust surface is distributed, incomplete, or still evolving.
Observed surfaces
Homepage; security docs; certification or compliance-related public language; privacy; terms; careers; real-time context/retrieval/ranking engine observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Zilliz Cloud

AI database / vector data layer

partial public trust surface
Evidence pattern
Official evidence present, but trust surface is distributed, incomplete, or still evolving.
Observed surfaces
Homepage; Trust Center navigation; Zilliz Cloud; Milvus; vector lakebase/vector database positioning; privacy/terms candidate URLs; careers observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Sparse / gap examples

Standalone vendors or serious project entities with verified official evidence, but a sparse public trust surface during review.

Cognee

AI database / vector data layer

sparse or gap pattern
Evidence pattern
Serious standalone entity with official evidence, but public trust surface remains sparse or needs profile caveats.
Observed surfaces
Homepage; docs; GitHub; Cognee Cloud; enterprise tier; privacy policy; US/EU terms; careers; open-source agent memory platform positioning observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Evidently AI

AI evaluation / observability

sparse or gap pattern
Evidence pattern
Serious standalone entity with official evidence, but public trust surface remains sparse or needs profile caveats.
Observed surfaces
Homepage; open-source AI evaluation/observability; privacy; terms; docs/GitHub observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Latitude

AI evaluation / observability

sparse or gap pattern
Evidence pattern
Serious standalone entity with official evidence, but public trust surface remains sparse or needs profile caveats.
Observed surfaces
Likely official site candidate for Latitude AI/evaluation workflow observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

LiteLLM

AI infrastructure / platform

sparse or gap pattern
Evidence pattern
Serious standalone entity with official evidence, but public trust surface remains sparse or needs profile caveats.
Observed surfaces
Homepage; docs; AI gateway; enterprise self-hosted/air-gapped, JWT auth, SSO, audit logs; careers observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Apify

AI infrastructure / platform

sparse or gap pattern
Evidence pattern
Serious standalone entity with official evidence, but public trust surface remains sparse or needs profile caveats.
Observed surfaces
Homepage; security; privacy; terms; marketplace/tools for AI and web automation; developer and enterprise automation surface observed during review.
What not to infer
Do not infer active buying intent, current procurement pressure, compliance quality, security quality, non-compliance, or absence of controls from observed or missing public surfaces.

Selected standalone examples

All rows below were included in the company profile candidate queue. This table is an index for the research report. Use the evidence matrix for horizontal comparison across observed public evidence surfaces.

Company Category role Maturity pattern Last checked
LangChain AI evaluation / observability benchmark evidence mature 2026-07-09
Baseten model deployment / fine-tuning benchmark evidence mature 2026-07-09
OpenAI foundation model / model provider benchmark evidence mature 2026-07-09
Pinecone AI database / vector data layer benchmark evidence mature 2026-07-09
Langfuse AI evaluation / observability benchmark evidence mature 2026-07-09
Weights & Biases AI evaluation / observability benchmark evidence mature 2026-07-09
n8n AI infrastructure / platform benchmark evidence mature 2026-07-09
Bland AI AI infrastructure / platform benchmark evidence mature 2026-07-09
Comet AI evaluation / observability benchmark evidence mature 2026-07-09
Groq AI infrastructure / platform benchmark evidence mature 2026-07-09
Hugging Face AI infrastructure / platform partial public trust surface 2026-07-09
Replicate model deployment / fine-tuning partial public trust surface 2026-07-09
Mistral AI foundation model / model provider partial public trust surface 2026-07-09
Not Diamond model deployment / fine-tuning partial public trust surface 2026-07-09
Ollama LLM developer tooling partial public trust surface 2026-07-09
Activeloop AI database / vector data layer partial public trust surface 2026-07-09
Firecrawl LLM developer tooling partial public trust surface 2026-07-09
Shaped AI database / vector data layer partial public trust surface 2026-07-09
Zilliz Cloud AI database / vector data layer partial public trust surface 2026-07-09
Cognee AI database / vector data layer sparse or gap pattern 2026-07-09
Evidently AI AI evaluation / observability sparse or gap pattern 2026-07-09
Latitude AI evaluation / observability sparse or gap pattern 2026-07-09
LiteLLM AI infrastructure / platform sparse or gap pattern 2026-07-09
Apify AI infrastructure / platform sparse or gap pattern 2026-07-09
Weaviate AI database / vector data layer benchmark evidence mature 2026-07-09
Modal AI cloud compute / model execution benchmark evidence mature 2026-07-09
Fireworks AI model serving / inference infrastructure benchmark evidence mature 2026-07-09
Scale AI AI data infrastructure / model services benchmark evidence mature 2026-07-09
Anthropic foundation model / model provider benchmark evidence mature 2026-07-09
Cohere foundation model / model provider benchmark evidence mature 2026-07-09
Supabase AI data infrastructure / developer data platform benchmark evidence mature 2026-07-09
Vercel AI app deployment / developer cloud platform benchmark evidence mature 2026-07-09