Inference provider uptime rankings
Uptime and speed of inference providers serving the same open models. Every 5 minutes we ask each one for a single token from gpt-oss-120b or GLM-5.3, from five regions, with our own key.
Uptime this week
Week 2026-W47 so far, as of Sat 21, 07:46 UTC. Live and not final: the ranking freezes on Monday at 00:05 UTC. Updated every 5 minutes.
- –Cerebrassince Thu 19, 12:00 UTC100%
- –Fireworks AIsince Thu 19, 12:00 UTC100%
- –Groqsince Thu 19, 12:00 UTC100%
- –Together AIsince Thu 19, 12:00 UTC100%
- –Novita AIsince Thu 19, 12:00 UTC100%
- –OpenRoutersince Thu 19, 12:00 UTC100%
- –Nebius AI Studiosince Thu 19, 12:00 UTC100%
- –SambaNovasince Thu 19, 12:00 UTC100%
- –DeepInfrasince Thu 19, 12:00 UTC92.67%
- –Baseten Model APIsNo results yet–
- Answered (hover a day for its uptime)
- Confirmed outage: 2+ regions failed on 2 runs in a row
- One run failed (shown, not counted)
- Not run yet
Probes run every 5 minutes from San Francisco, Ashburn, Montreal, Frankfurt and Singapore. Rank numbers appear once a provider has results for 90% of the week; until then the order is uptime so far.
What we test
| Provider | Model | API host | Links |
|---|---|---|---|
| Groq | gpt-oss-120b | api.groq.com | Status page · Is it down? |
| Together AI | gpt-oss-120b | api.together.xyz | Status page · Is it down? |
| Cerebras | gpt-oss-120b | api.cerebras.ai | Status page |
| DeepInfra | gpt-oss-120b | api.deepinfra.com | – |
| SambaNova | gpt-oss-120b | api.sambanova.ai | Status page |
| Nebius AI Studio | gpt-oss-120b | api.studio.nebius.com | Status page |
| Baseten Model APIs | gpt-oss-120b | inference.baseten.co | Status page · Is it down? |
| Fireworks AI | GLM-5.3 | api.fireworks.ai | Status page · Is it down? |
| Novita AI | GLM-5.3 | api.novita.ai | Status page |
| OpenRouter | gpt-oss-120b (OpenRouter routing) | openrouter.ai | Status page · Is it down? |
How we measure
- Request
- A 1-token chat completion with our own API key
- How often
- Every 5 min from 5 regions: San Francisco, Ashburn, Montreal, Frankfurt, Singapore
- Counted as down
- At least 2 regions fail on 2 runs in a row. Why confirm
- Not counted
- Rate limits (429), errors with our own key or billing, and outages of our own probes
- Speed
- Request to reply, minus DNS, connect and TLS. Compared only on the same model
- Week
- Monday to Sunday UTC, frozen when it ends. Ranked with 90% of the week measured
FAQ
Why does this differ from the vendor status page?
We test one API endpoint with one small request; vendors report on many products and regions with their own rules, so both can be right.
Which model do you call?
For LLM APIs, each provider's lowest-cost model. Inference providers all serve gpt-oss-120b or GLM-5.3, so their speed is comparable.
Can a short outage be missed?
Yes. Confirmation takes two 5-minute runs, so an outage shorter than about 10 minutes may not count.
Monitor your own model API with the same check: your key, your model, alerts when it fails.
Start free