Status: pre-launch. This page describes target design; this document does not establish a deployed service.
Updated sequencing — 2026-09-26. Two proposed sequences are kept apart. User products: voice-first messenger → RELAY broadcasts and milestones / CDN → inference. Node services: consensus participation → optional VPN / CDN → inference hosting. The tier table below is an earlier node-service proposal: a historical candidate for staging service research, not a required progression for every node. An operator may qualify separately for ledger participation, transport, optional third-party VPN egress, RELAY/CDN or inference. No listed service, reward, market size or pricing rule is established as live or verified here. See Communication Mesh Research.
Tiered Application Path — VPN / CDN / Livestream / Inference
This page describes the intended sequencing of a future SPECTRE commercial lane. Read product, gate, and revenue language below as target strategy unless a narrower runtime proof has explicitly closed the lane.
The Strategy
SPECTRE does not start as a full AI compute network, and it should not displace the rooted Skyzai company sequence. If SPECTRE becomes a live commercial lane, the proposal is to start with consensus participation, then optional VPN and CDN service, then inference hosting. Trust should be earned before complexity, with separate readiness gates for each optional service; operators need not qualify in one universal sequence.
Current company root: NEXUS -> Private Business Account -> API PAY -> proof event -> POS
Earlier node-service proposal:
Tier 1: VPN ──gate──> Tier 2: CDN ──gate──> Tier 3: Livestream ──gate──> Tier 4: Inference
Revenue can become cumulative later, but the first strategic question is sequence, not immediate yield.
Tier 1: Encrypted Relay (VPN)
Market: $50B global VPN market, growing 15% annually What it is: In the target design, SPECTRE nodes would relay encrypted traffic between endpoints, like a VPN run by many independent operators rather than one provider. Each operator would still see the addresses it connects and could keep logs; the design does not yet prevent that. Proof of Relay is a proposed incentive, not a working mechanism.
Requirements:
- Run a SPECTRE node
- Minimum bandwidth: 10 Mbps symmetric
- Stake: Minimum tier entry stake
- Uptime: 95%+ over rolling 7-day window
Revenue: Relay fees or subscription economics once the network lane is real.
Why it works first inside SPECTRE: VPN is the simplest buyer story and the lowest-operational-burden path. It is still not the company's first public GTM surface overall.
Gate to Tier 2: 30 days of 98%+ uptime with P-score > 0.7.
Tier 2: Content Delivery (CDN)
Market: $25B global CDN market What it is: In the target design, Tier 2 nodes would cache and serve static content (images, video, web assets) using SPECTRE's myelinated routing for optimal path selection. RELAY would broadcast push content to edge nodes. SPECTRE's topology map would help requests route to the nearest, fastest cache.
Requirements (in addition to Tier 1):
- Storage: Minimum 500GB SSD dedicated to CDN cache
- Bandwidth: 50 Mbps symmetric
- P-score: > 0.7 (proven honest behavior from Tier 1)
- Geographic diversity: Bonus rewards for underserved regions
Revenue: Per-GB delivery fees (higher than VPN due to caching value) plus topology map data licensing share.
What Tier 2 adds over Tier 1:
- Content caching and edge serving
- RELAY integration for push-based content distribution
- Participation in topology map data product
- Higher per-GB rates reflecting CDN premium
Gate to Tier 3: 60 days at Tier 2 with 99%+ uptime, P-score > 0.8, demonstrated storage reliability.
Tier 3: Livestream / Real-Time Media
Market: live events, creator streams, premium real-time distribution What it is: A specialized application layer for real-time video and audio distribution on top of proven routing and delivery infrastructure.
Requirements (in addition to Tier 1 + Tier 2):
- Stable ingest and egress handling
- Stricter latency targets than CDN
- Abuse, moderation, and support readiness
- Proven edge reliability under live traffic
Why it is not first in this proposal: Livestream has higher operational and reputational complexity than VPN or CDN. A live failure is public and immediate. It should sit on top of working distribution rather than serving as the first proof of that distribution.
Gate to Tier 4: Demonstrated live-delivery reliability at acceptable latency and support load.
Tier 4: Distributed Inference
Market: $150B+ AI infrastructure market What it is: In the target design, Tier 4 nodes would contribute GPU compute to distributed AI inference. SPECTRE's routing intelligence would direct inference requests to the optimal node based on model availability, latency, cost, and current load -- all computed by SmallEBM in real-time.
Requirements (in addition to Tier 1 + Tier 2 + Tier 3 where applicable):
- GPU: Minimum NVIDIA RTX 3060 equivalent (8GB VRAM)
- Bandwidth: 100 Mbps symmetric
- P-score: > 0.8 (proven reliable from Tier 1 + Tier 2)
- Model support: At least one supported inference model loaded
Revenue: Per-token inference fees (significantly higher than VPN/CDN) plus all lower-tier revenue streams.
What Tier 4 adds:
- GPU inference serving for LLMs, image models, embedding models
- Model routing via SmallEBM (latency-optimized model selection)
- Participation in the organism's Rosetta pipeline (serving as compute for Circle, ReFu, or APU agents)
- Premium topology map data (GPU availability and inference latency data)
The Gate Mechanism
In the target design, gates are not administrative approvals. They are algorithmic thresholds:
| Gate | Metric | Threshold | Measured By |
|---|---|---|---|
| 1 -> 2 | Uptime | 98%+ for 30 days | Consensus metadata |
| 1 -> 2 | P-score | > 0.7 | Immune system |
| 2 -> 3 | Uptime | 99%+ for 60 days | Consensus metadata |
| 2 -> 3 | P-score | > 0.8 | Immune system |
| 2 -> 3 | Storage reliability | < 0.01% cache miss on served content | CDN audit |
| 3 -> 4 | Live delivery reliability | threshold set by stream SLA | media audit |
No human would approve tier promotion. The network's own measurements would determine eligibility. In the target design, this prevents gatekeeping and ensures that only proven-reliable nodes handle higher-value traffic.
Strategic Rule
SPECTRE is a later lane.
Do not let VPN, CDN, livestream, or inference displace the current root company sequence:
NEXUS -> Private Business Account -> API PAY -> proof event -> POS
When the company has earned a SPECTRE lane, the proposed node-service order is
consensus participation -> optional VPN / CDN -> inference hosting; livestream
would sit within the CDN step. The order is a proposal, not a settled plan.