DownForAI

Ollama status: API, auth, latency & outage reports

Operational
Last probe 17 min ago·1 surface
Likely your sideMEDIUM confidence
📡 Official status page →📚 Docs →💳 Pricing →
100.0%
24h Uptime
76ms
p50 Latency
141ms
p95 Latency
0
Incidents (30d)

Having issues with Ollama?

Report problems quickly and help the community stay informed.

0
reports in the last 24 hours

Surface Health

Ollama AppOperational
HTTP 200p50 76ms16m ago

Uptime — last 24h

100.0%

Latency — last 24h (p50 per 30 min slot)

Loading…

Is Ollama down for everyone?

Likely local or client-side issue
Our probes see normal responses. The issue is likely on your end or in your network path.
Moderate confidence
Probe summary (1 surface)
All surfaces operational as of last probe.
Signals detected
  • All monitored surfaces operational
  • No recent user reports
  • Check your network, credentials, or rate limits

Incident history (30d)

✓No incidents recorded in the past 30 days.

Reported symptoms

No user reports for Ollama in the last 24 hours.

Known error signatures

Common failure patterns and how to diagnose them

Provider details

Popular local LLM runtime with local API, CLI, and cloud-hosted models. Official docs explicitly document integrations with tools such as Claude Code and Codex via ollama launch.

What we monitor
Ollama Registry — Critical for new model pulls
Ollama Cloud API — Managed inference (public beta since Jan 2026)
Ollama website — Docs, downloads
Status page segments
CloudRegistryWeb
Model families
User pulls what they want: Llama 3.3, Qwen 2.5, Mistral, Phi-3, DeepSeek R1, Gemma, etc.Also supports vision models (LLaVA, Llama 3.2 Vision) and embedding models
Common limits & quotas
  • Local deployment: unlimited, bounded only by user hardware (RAM, VRAM, disk)
  • Ollama Cloud: authenticated via API key, tiered plans (Free / Pro / Max); usage limits reset hourly and weekly per official pricing
  • No auth required for localhost API; auth required for Cloud API
Ecosystem dependencies
Integrates with: Claude Code, Codex CLI, Continue.dev, Zed, Cursor, Raycast AI, Open WebUI, LobeChat, LangChain, LlamaIndexOllama docs now explicitly support launching external integrations like Claude Code and Codex via ollama launchModels pulled from: Ollama Registry + Hugging Face (direct URL)Runs on: macOS, Linux, Windows, Docker
Operator notes
  • Key distinction: local Ollama and Ollama Cloud are different infrastructures. Local can work fine while Cloud is down, and vice versa.
  • Auth difference: localhost API requires no authentication; Cloud API requires API key. Never expose localhost:11434 to the public internet without auth middleware.
  • Check ollama --version and compare with the latest release if you suspect daemon/model compatibility issues
  • Many 'Ollama is down' reports are actually local daemon or hardware issues rather than platform outages
  • When downforai shows Ollama 'operational', we monitor the Cloud registry and API; local runs are the user's responsibility
Diagnostic signals
$ curl http://localhost:11434/api/tags — local daemon reachability (should return installed models list)$ curl http://localhost:11434/api/chat -d '{"model":"llama3.3","messages":[{"role":"user","content":"ping"}]}' — real local inference test$ ollama ps — active model processes$ ollama ls — list installed models$ ollama serve — start daemon manually$ ollama pull <model> — test Registry connectivity by pulling a small model

Fallback alternatives

What to use if this service is down

Registry is down but you need a cached model
Local ollama run <cached-model> continues to work
Easy switch
Registry down and need new model
Hugging Face direct download + Modelfile import
Moderate effort
Ollama Cloud degraded
Groq, Together AI, Fireworks AI host many of the same open-weight models via OpenAI-compatible APIs
Easy switch
Local daemon unstable on current hardware
LM Studio (GUI) is a drop-in alternative for the same models
Easy switch
Local daemon is healthy but model pull fails
Test already-cached local models before switching provider
Easy switch

How we monitor

downforai.com probes each AI service every 2–5 minutes from multiple independent locations. We measure HTTP response codes, latency (p50 & p95), and endpoint availability across the surfaces listed above. Status is classified as Operational, Degraded, or Outage based on a weighted combination of probe results. Uptime is calculated over 30-minute buckets and the last 24 hours. User reports are factored into our diagnosis as a secondary signal. We are independent of all providers listed and receive no compensation to report any particular status.

Community Discussion

0/1000
No comments yet. Be the first to share your experience!