Status
Bee status
Live snapshot of the Bee Intelligence Engine. The dashboard polls /api/health every 15 seconds and reports backend liveness and round-trip latency. The per-tier and per-integration rows below currently share that single backend verdict — independent per-component probes land with the observability roadmap milestone. For incident history beyond the rolling check, see the changelog.
Checking…
Awaiting backend response
Inference tiers
Cell
google/gemma-4-E4B-it on Modal CPU
Brood
Modal CPU, scale-to-zero
Comb
Modal L40S GPU
Buzz
Modal A100 80GB GPU
Hive
Modal H200 GPU
Swarm
Vercel fan-out router
Enclave
On-prem / customer-managed
Tiers currently share the parent backend probe. Per-tier health endpoints land with the observability roadmap milestone.
Platform & integrations
Bee API (chat completions)
Audio — Whisper STT (Groq)
Audio — Orpheus TTS (Groq)
Vision (NVIDIA Build)
RAG embeddings (NVIDIA NV-Embed)
RAG reranker (NVIDIA Mistral 4B)
Prompt-injection guard (Llama Prompt Guard)
Content safety (Llama Guard 4)
Billing (Stripe)
These integrations are not yet probed independently — each row reflects the overall Bee API liveness verdict above, not a per-provider health check. Per-integration probes land with the observability roadmap milestone.
SLA & incident notifications
Production availability targets are described per plan on the pricing page. SEV-1 incidents are notified to the account-owner email and posted as a changelog entry within 14 days; we do not sub-process incident details to a third-party status-page vendor.