◎
Same claim boundary
Shadow-mode advisory analytics on every tier. No ECU writes. Not ISO 26262 / ASIL safety-certified. Not an OEM homologation replacement.
Algorithm tiers · June 2026
Shadow-mode on every tier
Per-vehicle · API · Fleet
NADIR perception intelligence
Choose your depth of insight — not your trust boundary.
Like LLM providers offer Opus, Sonnet, and Haiku, NADIR offers four algorithm tiers — Apex, Meridian, Pulse, and Echo — each mapping to different math depth, latency, and price. Every tier runs in shadow mode with the same claim boundary. You pick how much forensic depth your fleet needs.
Why tiered algorithms
Variable pricing without variable trust
Not every fleet needs a 14-day stochastic forecast and CRLB ellipsoid on day one. Tiering lets telematics partners embed lightweight drift gates at scale, while OEM programs get full forensic depth — without changing what NADIR promises (or doesn't promise) about safety certification.
↗
Natural upsell path
Start with Pulse for volume compliance. Graduate to Meridian as maintenance workflows mature. Upgrade to Apex when insurers or OEM auditors demand full evidence graphs.
⚡
API-native selection
Pass ?tier=meridian on scoring routes, or set an org-level feature flag. Scopes: drift:pulse vs horizon:forecast gate compute-heavy endpoints.
Deep dive
Every tier, in full detail
Who each tier serves, what math runs under the hood, and exactly what is — and isn't — included.
NADIR Apex
Premium · highest quality · full forensic depth
Mission critical
Latency: 2–8s / vehicle
HORIZON: full 6 modules
HORIZON: full 6 modules
Who this serves
- OEM ADAS program offices running homologation-adjacent field monitoring
- Large commercial fleets (Amazon-class last-mile, FedEx/UPS regional) with dedicated safety engineering
- Insurer-backed fleet programs requiring chain-of-custody evidence at claim close
- MSO networks (Caliber, Safelite) needing audit-grade post-repair QA at scale
- Fleet safety directors who buy on "which vehicles go CRITICAL in 14 days?"
What Apex supports
- Full stochastic 14-day forecast (algorithms A1–A6)
- BOCPD + HSMM + spectral GLR changepoint ensemble
- CRLB identifiability ellipsoids per sensor frame (3D twin)
- SE(3) geodesic drift replay and factor-graph evidence theater
- Fleet prioritizer with budget-constrained knapsack intervention planner
- What-if counterfactual forecasting (recalibrate / scenario coupling)
- Signed evidence bundles with full factor-graph export
- WebSocket live incident rail + geospatial identifiability map
- Unlimited evidence exports · SLA-backed support
What Apex does not include
- On-vehicle ECU writes or in-loop ADAS control
- OEM homologation sign-off or ISO 26262 certification
- Insurer underwriting certification (evidence is advisory)
- Guaranteed prediction accuracy — forecasts are operational analytics
- Real-time edge inference on the vehicle (cloud/batch compute)
NADIR Meridian
Balanced · efficient · production default
Default pilot SKU
Latency: 0.5–2s / vehicle
Best margin vs compute
Best margin vs compute
Who this serves
- Regional last-mile operators (50–500 vehicles) — food distribution, DSP contractors, shuttle fleets
- Fleet safety managers who need a CRITICAL queue and weekly digest without a data science team
- Maintenance program sponsors evaluating shadow pilots ($750 micro → $750/mo conversion path)
- Repair networks running post-windshield or post-collision calibration QA
- Buyers who want HORIZON Command Deck + Forecast Studio without full 3D twin overhead
What Meridian supports
- Continuous drift tier classification (NOMINAL / CAUTION / CRITICAL)
- Linear + accelerating regime forecast with trust-region gating
- Batch residual scoring with idempotency keys
- Time-to-tier survival curves (Weibull MLE, t10/t50/t90)
- Console dashboard · 3 seats included · weekly PDF + JSON digest
- CRITICAL vehicle queue for maintenance scheduling
- Signed evidence bundle samples (limited exports on subscription)
- Fleet health summary API + CSV export
- SSE residual score stream for live Console updates
What Meridian does not include
- Full BOCPD + HSMM + spectral GLR ensemble (single changepoint lane only)
- CRLB 3D ellipsoid rendering or Fleet Twin WebGL module
- Budget-optimized fleet knapsack prioritizer (ranked list only)
- Factor-graph 3D evidence theater or geodesic SE(3) replay scrubber
- What-if counterfactual intervention planner
- Geospatial route identifiability heatmap
NADIR Pulse
Fast · real-time edge · high-volume
Volume & embed
Latency: <200ms / vehicle
Telematics partner SKU
Telematics partner SKU
Who this serves
- Telematics platform vendors embedding a "sensor reliability" module
- High-volume fleets (1,000+ vehicles) needing tier gates, not forecasts
- Compliance-driven operators who must prove continuous monitoring exists
- Edge deployments where bandwidth and compute budgets are tight
- Partners who white-label NADIR scoring inside their existing dashboard
What Pulse supports
- Basic drift detection from existing telemetry (no new hardware)
- Lightweight Mahalanobis distance scoring
- CAUTION and CRITICAL tier gates with confidence bands
- Batch telemetry ingest API (
telemetry:ingestscope) - Webhook alerts on tier transitions
- JSON fleet health summary (tier counts, no survival curves)
- Rate-limit-friendly — optimized for 10k+ vehicles / org
What Pulse does not include
- 14-day survival forecast or time-to-CRITICAL probabilities
- 3D digital twin, CRLB ellipsoids, or HORIZON SPA modules
- Changepoint detection (BOCPD, HSMM, spectral GLR)
- Signed evidence bundle exports (webhook JSON only)
- Console seats or weekly PDF digest (Meridian+) — Pulse micro pilot includes simplified console + digest
- Fleet prioritizer, what-if planner, or intervention recommendations
- SE(3) extrinsic history drill-down (current snapshot only)
NADIR Echo
Optional future · batch · archival replay
Audit & archive
Cadence: weekly batch
No live WebSocket
No live WebSocket
Who this serves
- Legal and compliance teams needing periodic audit trails, not real-time ops
- Insurance carriers reviewing fleet calibration history after incidents
- Annual OEM field-reliability reports from archived telemetry
- Fleets that already run Meridian live but want quarterly forensic replay at Apex depth
- Post-pilot archives — prove what NADIR detected during evaluation window
What Echo supports
- Weekly (or on-demand) batch replay of full fleet telemetry
- Evidence export packages in signed ZIP bundles
- Historical tier timeline reconstruction
- Replay at Meridian or Apex math depth (buyer's choice per export)
- CSV + JSON summaries for counsel and auditors
- No live stream infrastructure required — lowest ongoing infra cost
What Echo does not include
- Live WebSocket incident rail or real-time Console updates
- Maintenance queue or dispatch integration
- Continuous monitoring SLA (batch cadence only)
- HORIZON interactive modules (static export artifacts only)
- Telemetry ingest API (replay consumes stored archives only)
Capability matrix
Compare tiers at a glance
| Capability | Apex | Meridian | Pulse | Echo |
|---|---|---|---|---|
| Tier classification (NOMINAL/CAUTION/CRITICAL) | ✓ | ✓ | ✓ | ✓ (replay) |
| Mahalanobis residual scoring | ✓ full | ✓ | ✓ lightweight | ✓ |
| 14-day stochastic forecast (A1) | ✓ | Regime only | — | On export |
| Survival curves / time-to-CRITICAL (A2) | ✓ | ✓ | — | On export |
| CRLB 3D ellipsoids (A3) | ✓ | — | — | Apex export |
| Fleet knapsack prioritizer (A4) | ✓ | Ranked list | — | — |
| SE(3) geodesic replay (A5) | ✓ | — | — | Apex export |
| Factor-graph evidence (A6) | ✓ | — | — | ✓ |
| Changepoint ensemble (BOCPD/HSMM/GLR) | ✓ | Single lane | — | Apex export |
| HORIZON Command Deck | ✓ | ✓ | — | — |
| HORIZON 3D Fleet Twin | ✓ | — | — | — |
| HORIZON Map + identifiability | ✓ | — | — | — |
| Console dashboard | ✓ unlimited | ✓ 3 seats | ✓ Pulse view | — |
| Weekly PDF digest | ✓ | ✓ | ✓ | On export |
| Signed evidence bundles | ✓ unlimited | Limited | — | ✓ per export |
| Live WebSocket stream | ✓ | ✓ SSE | Webhooks | — |
| What-if intervention planner | ✓ | — | — | — |
| Typical latency per vehicle | 2–8s | 0.5–2s | <200ms | Batch (weekly) |
Algorithm tier pricing
Per-vehicle rates by tier
All prices USD · billed monthly unless noted · 10% discount on annual prepay. Shadow-mode advisory analytics on every tier — binding commercial terms in LOI exhibit or MSA.
Pulse
$3–5
per vehicle / month
- Mahalanobis drift + tier gates
- Pulse console + weekly PDF digest
- CRITICAL queue + webhooks
- Telemetry ingest API
- Calibration Lab Pulse parity demo
Micro pilot: $750 · 4 weeks · 50–200 vehicles
Pulse micro pilot →Meridian
$3.75
per vehicle / month
- Everything in Pulse
- Survival curves + regime forecast
- Console (3 seats) + weekly digest
- CRITICAL maintenance queue
- 5 signed evidence exports / mo
- HORIZON Command Deck + Forecast Studio
Min 50 vehicles · $750/mo micro-fleet cap at ≤200
Apex
$8.50
per vehicle / month
- Everything in Meridian
- Full A1–A6 forecast ensemble
- HORIZON all 6 modules (3D twin, map, theater)
- Unlimited evidence exports
- Budget optimizer + what-if planner
- SLA-backed support · dedicated FDE
Min 100 vehicles · enterprise quote above 2,000
Echo
$0.40
per vehicle / export
- Weekly batch fleet replay
- Signed evidence ZIP per run
- Choose Meridian or Apex replay depth
- CSV + JSON auditor summaries
- No live infra · lowest ongoing cost
Or $12k/yr audit bundle (unlimited exports, ≤500 veh)
Claim boundary (all tiers): NADIR is shadow-mode reliability analytics and forecasting. It is not ISO 26262 / ASIL safety-certified, not an OEM homologation replacement, and must not be used for in-loop vehicle control. All forecasts and tier classifications are advisory operational analytics. Indicative pricing — binding terms in separate commercial agreement.
Developer & API access
API pricing — build on NADIR without the Console
Per-vehicle tier pricing covers algorithm depth. API access is additive — for integrators, telematics embeds, and custom workflows. All routes org-scoped with Bearer API keys and RBAC scopes.
API platform tiers
Requires an active algorithm tier (Pulse minimum for ingest)
Developer sandbox
Free · 12-vehicle sample fleet · all scopes for integration testing
$0
Evidence API add-on
GET /v1/evidence/{id}/bundle · signed JSON + chain of custody · batch ZIP export
+$200/mo
HORIZON forecast API
horizon:forecast scope · /v1/horizon/forecast/* · fleet-forecast + twin payloads
+$350/mo
Full platform API
Telemetry ingest + webhooks + fleet export + evidence + horizon · included at Apex; add-on for Meridian
+$500/mo
Webhook delivery worker
Async retry + backoff · tier transition · evidence-ready events
+$150/mo
Extra console seats
5-pack · Meridian and above
+$150/mo
API rate limits & quotas
Per-organization · sliding window · upgradeable
Pulse — ingest bucket
Batch telemetry POST · 50k requests/day
Included
Meridian — read + export
Fleet health, residual score, CSV export · 20k requests/day
Included
Apex — forecast compute
/v1/horizon/* · fleet-forecast cached 60s · 5k compute calls/day
Included
Overage — ingest
Additional 10k batch ingest requests
$50/block
Overage — evidence export
Additional 100 signed bundle downloads
$75/block
Dedicated API key rotation + audit
Overlapping grace period · append-only audit log
Enterprise
API scopes: fleet:read fleet:write telemetry:ingest evidence:read horizon:read horizon:forecast drift:pulse platform:* —
OpenAPI contract at NADIR_SDK/openapi/nadir.v1.contract.json (67 operations).
Fleet analytics packages
Pilot programs & fleet subscriptions
Algorithm tier pricing is per-vehicle. Fleet packages bundle onboarding, support, and surface access for common buyer journeys — from $750 micro pilots to enterprise fleet subscriptions.
Micro · Pulse
Pulse micro-fleet pilot
$750 one-time
50–200 vehicles · 4 weeks · Pulse algorithm tier. Tier gates + digest — no HORIZON forecast.
- Pulse console + weekly PDF digest
- CRITICAL vehicle queue
- Calibration Lab Pulse parity walkthrough
- Business-hours email support
- Converts to $3–5/veh/mo Pulse subscription
Paid evaluation
Fleet safety pilot
$2,800 6 weeks
100–500 vehicles · Meridian tier · maintenance or safety budget holder.
- Everything in shadow evaluation
- Signed evidence samples for legal review
- Dedicated kickoff + close-out calls
- Written success metrics vs baseline
- Commercial proposal at close
Enterprise
Fleet subscription
From $4,500/mo
500+ vehicles · Apex or Meridian · annual contract · SLA-backed.
- Full telemetry ingest API + webhooks
- Unlimited signed evidence exports (Apex)
- HORIZON Command Deck + modules per tier
- Custom metrics dashboard + success criteria
- Dedicated forward-deployed engineer (Apex)
Free shadow evaluations
Qualified design partners · 4–6 weeks · Meridian tier
HORIZON design partner
Command Deck + 14-day forecast walkthrough · 80–300 vehicles
$0
Enterprise LOI cohort
Amazon Logistics, Safelite, Caliber-scale · signed exhibit required
$0
Conversion path
Paid pilot ($2,800) or fleet subscription upon success criteria
On exit
HORIZON analytics surfaces
Which modules unlock per algorithm tier
Command Deck + Forecast Studio
Meridian and above · live KPI rail + survival curves
Meridian+
3D Fleet Twin + CRLB ellipsoids
WebGL sensor frames · drift arcs · FOV volumes
Apex only
Horizon Map + route identifiability
deck.gl geospatial risk · regional heat
Apex only
Evidence Theater + What-If Console
Factor-graph 3D · intervention planner
Apex only
FAQ
Common pricing questions
Can I mix tiers across my fleet?
Yes. Many telematics partners run Pulse on the full fleet for tier gates, then apply Meridian or Apex to a CRITICAL subset or high-value routes. Billing is per-vehicle per active tier — your org dashboard shows the tier badge per vehicle group.
Is shadow mode really free for design partners?
Yes, for qualified enterprise design partners with a signed LOI exhibit (Amazon Logistics, Safelite, Caliber tier). NADIR absorbs compute during evaluation; conversion to paid tiers funds ongoing monitoring. HORIZON shadow runs 4–6 weeks at Meridian depth.
Why is Meridian the default pilot SKU?
Meridian delivers the best margin vs compute: survival curves and CRITICAL queue without the WebGL/forecast ensemble cost of Apex. The $750 micro-fleet pilot maps to ≤200 vehicles at Meridian — about $3.75/vehicle/mo at the cap.
Do I need the API add-ons if I'm on Console only?
No. Console access is included in Meridian and Apex per-vehicle pricing. API add-ons are for integrators embedding NADIR in telematics dashboards, insurer workflows, or custom maintenance systems. Evidence API (+$200/mo) is the most common add-on for MSO pilots.
What's the difference between Echo and a Meridian subscription?
Meridian is live continuous monitoring with weekly digests and real-time Console updates. Echo is batch-only: weekly or on-demand replay of archived telemetry with evidence export. Echo suits audit and legal review — not operations. You can run Meridian live and add Echo for quarterly forensic exports at Apex depth.
Is there an annual discount?
10% prepay discount on micro subscriptions, per-vehicle tier pricing, and fleet subscriptions. Echo annual audit bundle ($12k/yr for ≤500 vehicles) includes unlimited exports.
Pick a tier. Run shadow. Prove value.
Start with a $750 Pulse micro pilot, explore Meridian for survival curves, or request an Apex enterprise quote for 500+ vehicles.