SeedInfer Providers · Become a node

Gateway fleet /api/v1/providers · NVFP4 1M ctx $0.02/$0.05 · CUDA 13.3 · 47900/47901 · 0 nodes · last fetch —

Become a Provider →
NVFP4 · 1M ctxseedinfer/nemotron-lightning-1m$0.02 / $0.05 per 1MCUDA 13.3 · driver 580+47900:8000 + 47901:3001Docs / hardware ·Full guide on /provider →

One-liner — terminal Linux (plug-and-play)curl | bash

Runs: nvidia-smi check (VRAM 32GB min, 16GB hard min, ports 47900/47901 free → env VLLM_PORT/AGENT_PORT) → Docker + nvidia-ctk + tailscale → tailscale up --login-server https://tailnet.seedinfer.com --authkey XXX --advertise-tags tag:provider → docker compose up -d --build (host 47900:8000, 47901:3001) → heartbeat every 30s → verified fleet.

1 · Simple (recommended)
curl -fsSL https://seedinfer.com/install.sh | bash -s -- --authkey YOUR_AUTHKEY

Replace YOUR_AUTHKEY with a key from /api/v1/auth/request or the Generate button on /provider. Gateway: https://seedinfer.com.

2 · Auto-fetch key (jq)
curl -fsSL https://seedinfer.com/install.sh | bash -s -- --authkey $(curl -s https://seedinfer.com/api/v1/auth/request | jq -r .authkey)

For scripts — fetches authkey from /api/v1/auth/request and installs immediately. Requires jq.

3 · Full options + hostname
curl -fsSL https://seedinfer.com/install.sh | bash -s -- --authkey YOUR_AUTHKEY --model seedinfer/nemotron-lightning-1m --gateway https://seedinfer.com --hostname provider-5090
/install.sh·/api/install·/api/v1/auth/request·/docs → hardware

Quick requirements

ModelNVFP4 · 1M ctx $0.02/$0.05
GPU minRTX 5090 32GB GB202
VRAM math16-22 +6 KV =22-28GB
OSUbuntu 24.04 noble
Driver / CUDA580.65+ / 13.3
Docker24+ + nvidia-ctk
HF cache50GB+ (~60GB total)
Free space60GB+ free (df -h)
Ports47900:8000 + 47901:3001
NetUDP 41641
Flags host 1:1: marlin + flashinfer + fp8 · 0.93 · 1048576 · 128 · 4096 · VLLM_ATTENTION_BACKEND=FLASHINFER + tailscale auto.
Full hardware docs → /docs
Verify after install
curl -fsS http://127.0.0.1:47901/health | jq
curl -fsS http://127.0.0.1:47900/v1/models | jq
curl http://127.0.0.1:47901/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model":"seedinfer/nemotron-lightning-1m","messages":[{"role":"user","content":"ping"}],"max_tokens":32}'

Fleet: GET /api/v1/providers · heartbeat co 30s · auto-verify po 2 heartbeat.

Loading providers from /api/v1/providers + /api/stats …

SeedInfer gateway fleet — /api/v1/providers· primary

0 verified · 0 total
Add your node → ·JSON

Provider fleet · 0 total · showing 0

No providers match filters.
Fleet page /api/v1/providers returns verification.status + heartbeat_count + tailscale_ip. UI: pending🟡 / verifying🔵 opacity 60, verified🟢 opacity 100 (official), failed🔴. Manual: curl -X POST https://seedinfer.com/api/v1/providers/verify -H 'Content-Type: application/json' -d '{"provider_id":"xxx"}'
Upstream fleet — /api/stats (secondary, reference only)0 providers— active
SeedInfer Network Reference — data from /api/stats. SeedInfer gateway fleet (/api/v1/providers) is the source of truth — nodes with heartbeat + verification. Network stats serve only as a comparison. See /docs and /stats.

Provider fleet · 0 total · showing 0

No providers match filters.
SeedInfer.com · Gateway fleet /api/v1/providers (heartbeat + verification) + heartbeat POST /api/v1/providers/heartbeat · One-liner https://seedinfer.com/install.sh · Host 47900:8000 (vLLM) 47901:3001 (agent) · /docs · /provider · SeedInfer Network Statistics /api/stats (secondary collapsed).