AI Reliability Index
Live uptime and latency data for 48 major AI services. Updated hourly. Ranked by 30-day uptime.
| # | Service | Category | Uptime 24h | Uptime 7d | Uptime 30d | p50 | p95 | Incidents 30d |
|---|---|---|---|---|---|---|---|---|
| 1 | Google AI Studio | Dev Tools | 100.00% | 100.00% | 100.00% | 334ms | 637ms | 0 |
| 2 | Kling AI | Video | 100.00% | 100.00% | 100.00% | 733ms | 1083ms | 0 |
| 3 | ElevenLabs | Audio | 100.00% | 100.00% | 100.00% | 264ms | 431ms | 0 |
| 4 | Suno | Audio | 100.00% | 100.00% | 100.00% | 167ms | 412ms | 0 |
| 5 | GitHub Copilot | Dev Tools | 100.00% | 100.00% | 100.00% | 24ms | 399ms | 0 |
| 6 | Cursor | Dev Tools | 100.00% | 100.00% | 100.00% | 86ms | 310ms | 0 |
| 7 | Lovable | Dev Tools | 100.00% | 100.00% | 100.00% | 325ms | 618ms | 0 |
| 8 | Meta Llama | LLM | 100.00% | 100.00% | 100.00% | 630ms | 1127ms | 0 |
| 9 | Microsoft Copilot | LLM | 100.00% | 100.00% | 100.00% | 74ms | 388ms | 0 |
| 10 | OpenAI | LLM | 100.00% | 100.00% | 100.00% | 345ms | 691ms | 0 |
| 11 | Anthropic | LLM | 100.00% | 100.00% | 100.00% | 410ms | 777ms | 0 |
| 12 | Google Gemini | LLM | 100.00% | 100.00% | 100.00% | 147ms | 412ms | 0 |
| 13 | DeepSeek | LLM | 100.00% | 100.00% | 100.00% | 274ms | 481ms | 0 |
| 14 | Cohere | LLM | 100.00% | 100.00% | 100.00% | 92ms | 289ms | 0 |
| 15 | Hugging Face | Infrastructure | 100.00% | 100.00% | 100.00% | 76ms | 146ms | 0 |
| 16 | Groq | Infrastructure | 100.00% | 100.00% | 100.00% | 88ms | 237ms | 0 |
| 17 | Together AI | Infrastructure | 100.00% | 100.00% | 100.00% | 262ms | 458ms | 0 |
| 18 | Replicate | Infrastructure | 100.00% | 100.00% | 100.00% | 250ms | 370ms | 0 |
| 19 | Cerebras | Infrastructure | 100.00% | 100.00% | 100.00% | 401ms | 714ms | 0 |
| 20 | Adobe Firefly | Image | 100.00% | 100.00% | 100.00% | 91ms | 201ms | 0 |
| 21 | Krea AI | Image | 100.00% | 100.00% | 100.00% | 307ms | 784ms | 1 |
| 22 | Runway | Video | 100.00% | 100.00% | 100.00% | 85ms | 317ms | 0 |
| 23 | QuillBot | LLM | 100.00% | 100.00% | 100.00% | 657ms | 1018ms | 0 |
| 24 | LMArena | LLM | 100.00% | 100.00% | 100.00% | 483ms | 948ms | 0 |
| 25 | Ollama | LLM | 100.00% | 100.00% | 100.00% | 66ms | 137ms | 0 |
| 26 | Voicemod | Audio | 100.00% | 100.00% | 100.00% | 61ms | 427ms | 0 |
| 27 | Vast.ai | Infrastructure | 100.00% | 100.00% | 100.00% | 115ms | 234ms | 0 |
| 28 | Devin (Cognition) | Dev Tools | 100.00% | 100.00% | 100.00% | 86ms | 145ms | 0 |
| 29 | n8n | Agents | 100.00% | 100.00% | 100.00% | 78ms | 149ms | 0 |
| 30 | ChatGPT | LLM | 100.00% | 100.00% | 100.00% | 571ms | 2420ms | 0 |
| 31 | Mistral AI | LLM | 100.00% | 98.23% | 98.29% | 69ms | 250ms | 0 |
| 32 | Stability AI | Image | 100.00% | 98.21% | 98.28% | 114ms | 242ms | 0 |
| 33 | Moonshot AI (Kimi) | LLM | 94.12% | 98.18% | 98.25% | 1720ms | 3308ms | 0 |
| 34 | Candy AI | Roleplay AI | 92.86% | 98.08% | 98.15% | 122ms | 303ms | 0 |
| 35 | Tripo AI | 3D & Avatars | 94.44% | 92.98% | 93.22% | 1450ms | 5000ms | 0 |
| 36 | Baidu AI Cloud | Infrastructure | 88.24% | 78.57% | 77.59% | 3896ms | 5000ms | 0 |
| 37 | OpenAI Sora | Video | 100.00% | 42.86% | 41.38% | 168ms | 310ms | 1 |
| 38 | Grok Imagine | Image | 100.00% | 42.11% | 40.68% | 23ms | 420ms | 0 |
| 39 | Perplexity | Search | 100.00% | 41.59% | 40.17% | 55ms | 282ms | 0 |
| 40 | Character.AI | LLM | 100.00% | 41.23% | 39.83% | 259ms | 557ms | 0 |
| 41 | xAI Grok | LLM | 100.00% | 41.07% | 39.66% | 84ms | 213ms | 0 |
| 42 | Midjourney | Image | 100.00% | 41.07% | 39.66% | 36ms | 176ms | 0 |
| 43 | Viggle | Video | 100.00% | 41.07% | 39.66% | 195ms | 391ms | 0 |
| 44 | Poe | LLM | 100.00% | 40.54% | 39.13% | 185ms | 565ms | 0 |
| 45 | Genspark | LLM | 100.00% | 40.35% | 38.98% | 84ms | 461ms | 0 |
| 46 | Crushon AI | Roleplay AI | 100.00% | 40.35% | 38.98% | 383ms | 616ms | 0 |
| 47 | SillyTavern | Roleplay AI | 100.00% | 36.54% | 35.19% | 278ms | 622ms | 10 |
| 48 | Magnific AI | Image | 0.00% | 0.00% | 0.00% | 112ms | 225ms | 0 |
About this index: Data collected every 15 minutes from probes monitoring API endpoints. Uptime is the percentage of checks returning a healthy response. Latency p50/p95 computed from the last 30 days of observations. Free to use — attribution appreciated. See methodology for details.