Pricing and engagements
From strategy to runtime protection.
Starseer operates in three modes: the platform for continuous scanning, detection, and enforcement, advisory to sharpen your AI security strategy, and services for custom design and optimization. Most engagements start with the platform and expand from there.
Platform
Scan, detect, and enforce
Model vulnerability scanning, runtime detection and response, and gateway policy enforcement. Subscription tiers scoped to your environment.
Advisory
Risk and strategy advisory
A dedicated advisor working alongside your security and AI leadership on risk posture, strategy, and operational practice. Retained basis.
Services
Design and build
Model selection, LLM and RAG architecture, and optimization. Scoped project engagements for specialized needs.
Platform tiers
Continuous protection across the full lifecycle.
Three tiers scoped to how your team operates, spanning model vulnerability scanning before deployment, detection and response at runtime, and policy enforcement at the gateway. All tiers are custom-quoted based on environment scope.
Try and deploy
Essentials
A low-touch way to put runtime protection in front of a live endpoint, using a lightweight Canary model.
Detection and response
Jailbreak detection (Canary)
Lightweight model deployment
API integration
Free developer trial
Core support
Model vulnerability scan
Vulnerability scanning
Security-verified AIBOM generation
Model provenance verification
Automated report output
Full detection library
Advanced
Everything in Essentials, plus the full behavioral detection library and gateway enforcement.
All Essentials capabilities
Detection library
Destructive action detection
Data exfiltration patterns
Tool abuse and prompt injection
OWASP ASI coverage
Threat feed subscription
Gateway
Prompt classification
Model routing
Baseline policy enforcement
Right-sized spend
Custom rules and policy
Enterprise
Everything in Advanced, plus custom detections and enforcement policy written for your threat model.
All Advanced capabilities
Custom detection
Custom detections
Rule testing and validation
Organization-specific detections
Gateway policy
Custom policy enforcement
Advanced routing rules
Supply chain
Tamper and backdoor detection
Supply chain attestation
Compliance evidence
Support
Priority support
Add-on · analyst-assisted
Model Forensics
Our team runs the analysis, comparing behavior against known-good baselines and delivering findings with remediation guidance. For incident response or pre-deployment assurance.
Includes
Behavior comparison analysis
Known-good baseline diffing
Detailed findings report
Remediation guidance
Incident response report
Single-token backdoor scanning
Add-on · research
Labs
Deep interpretability work, model optimization, and custom research, run by the team that builds the platform. Pairs with any tier.
Includes
Fine-tuning optimization
Root cause analysis
Custom model tuning
Interpretability research
Dedicated advisory
All tiers are custom-quoted based on environment scope. Add-ons pair with any tier.
Request a demoAdvisory
Risk and strategy advisory.
A dedicated Starseer advisor works alongside your security and AI leadership on a retained basis, providing expert guidance on AI risk posture, security strategy, and operational best practices as your program matures.
Talk to an advisorPlatform deployment strategy and optimization guidance
AI risk posture assessment and security strategy development
Detection coverage tuning and operational best practices
Monthly strategic briefing with your CISO or AI security lead
Incident advisory and post-incident review
Priority access to Starseer research and threat briefings
Policy, governance, and compliance guidance
Assessment and services
Every services engagement starts with an assessment.
Before any design, build, or optimization work begins, we run the same four-phase methodology to map your environment, surface risks, and define priorities. Engagements are scoped per project or on a monthly retainer.
01 · Discover
Map your AI environment
Inventory all models, agents, pipelines, and data flows. Establish scope and identify the full AI attack surface before any analysis begins.
02 · Assess
Analyze models and exposures
Apply interpretability analysis and adversarial testing to surface backdoors, hidden capabilities, and behavioral risks. Map findings to MITRE ATLAS.
03 · Prioritize
Rank risks and close gaps
Score findings by business impact and exploitability. Identify detection coverage gaps and sequence remediation by risk level and operational feasibility.
04 · Deliver
Strategy, roadmap, and evidence
Deliver a security strategy report, prioritized remediation roadmap, and compliance-ready evidence package aligned to NIST AI RMF and ISO 42001.
Service engagements
Assessment-led, scoped to defined deliverables. Ongoing optimization retainers are available after initial delivery.
AI model selection
Evaluating which foundation models, fine-tuned variants, or specialized architectures fit your use case, accounting for capability, security posture, cost, compliance requirements, and deployment constraints. Prevents costly model lock-in.
AI system optimization
Performance and reliability optimization for AI systems already in production. Reduces token spend, improves latency, tightens accuracy, and resolves drift, hallucination, and reliability issues without weakening security posture.
AI, LLM, and RAG design
Architecture design for LLM-powered applications, retrieval-augmented generation pipelines, and agentic systems. Covers prompt architecture, retrieval design, guardrail strategy, data flow security, and deployment topology.
Scope, timeline, and fees are set during an initial discovery session.
Talk to servicesStart where your exposure is.
Four minutes with the diagnostic tells you which tier fits. A demo shows you the platform against your own workload.