{
 "metadata": {
  "name": "AI API provider SLA terms, as the providers state them",
  "readUtc": "2026-10-28",
  "method": "Each page was opened in Chrome on the read date and its rendered text read in full (Anthropic: the whole page was searched for uptime, service level, SLA, service credit and availability wording). Nothing here comes from a third-party summary or from memory. A provider with status 'not read' has no entry of terms: that is not the same as having no SLA.",
  "limits": [
    "Only publicly posted pages were read. Most SLAs sit in enterprise agreements that are not public; 'details specified in offline agreement' means the posted page is a summary.",
    "Posted terms can change at any time (Mistral reserves the right to update its schedule). Each row carries its read date.",
    "A stated uptime target is a commitment or an objective as worded by the provider; the 'kind' field keeps the provider's own word."
  ]
 },
 "providers": [
  {
   "id": "mistral", "name": "Mistral", "status": "read",
   "sources": [{ "url": "https://legal.mistral.ai/terms/enterprise-service-level-and-support-schedule/", "title": "Enterprise Service Level and Support Schedule", "effective": "2026-07-23", "readUtc": "2026-10-28" }],
   "kind": "Service Level Objective (the provider's term), with service credits",
   "applies_to": "Enterprise customers with an agreement; products AI Studio and Vibe, for the part deployed on Mistral infrastructure",
   "uptime_targets": [{ "tier": "Basic", "monthlyUptimePct": 99.5, "comparator": "at least" }, { "tier": "Standard", "monthlyUptimePct": 99.9 }, { "tier": "Premium", "monthlyUptimePct": 99.9 }],
   "downtime_definition": "Server-side downtime, counting only downtime periods of five minutes or more; Monthly Uptime Percentage is total minutes in the month minus those minutes, divided by total minutes.",
   "credits": "Basic: 2% of Monthly Fees at 99.4% to 95%, 5% below 95%. Standard: 2% at 99.8% to 99.5%, 5% at 99.4% to 95%, 10% below 95%. Premium: 5% at 99.8% to 99.5%, 10% at 99.4% to 95%, 20% below 95%. Total credits in a month capped at 25% of the amount owed for the affected product.",
   "claim": "Customer must email support-ent@mistral.ai within 30 days of the end of the calendar month, with the date and time of the failure; otherwise the right to a credit is forfeited. Credits are the sole and exclusive remedy and are applied within 60 days of the request (or reimbursed within 90 days of the end of the month if no further invoice).",
   "exclusions": "Alpha or beta features, third-party products, causes beyond reasonable control, customer software or infrastructure, customer breach, third-party modification, planned maintenance with advance notice.",
   "reporting": "Availability report at https://status.mistral.ai/",
   "change_clause": "Mistral may update the schedule at any time; a customer materially harmed can object within 30 days of notice and stays on the earlier schedule until the end of the order term or 12 months."
  },
  {
   "id": "openai", "name": "OpenAI", "status": "read",
   "sources": [{ "url": "https://openai.com/api-scale-tier/", "title": "Scale Tier for API Customers", "readUtc": "2026-10-28" }],
   "kind": "Uptime SLA and latency SLA (the provider's words) on Scale Tier",
   "applies_to": "Enterprise customers who buy Scale Tier token units (minimum 30 days each) for models released before GPT-5.6; the page points GPT-5.6 and later to Reserved Tier, which was not read",
   "uptime_targets": [{ "tier": "Scale Tier, every model listed", "monthlyUptimePct": null, "pct": 99.9, "note": "the page states 99.9% uptime SLA; it does not define the measurement period or what counts as downtime" }],
   "latency_sla": "99% of requests above a stated tokens-per-second figure per model (for example 50 for GPT-5.5, 100 for GPT-5.4 mini), calculated as p50 request latency per 5 minutes",
   "downtime_definition": "not stated on this page",
   "credits": "Amount not stated on this page. Where both the latency SLA and the uptime SLA are violated, the customer is credited the greater of the two SLA amounts for the calendar month of the token unit purchase.",
   "claim": "not stated on this page",
   "exclusions": "not stated on this page",
   "reporting": "not stated on this page",
   "pay_as_you_go": "The page states no uptime SLA for pay-as-you-go use; it points to Fast mode for the same latency and reliability benefits on a flexible basis."
  },
  {
   "id": "groq", "name": "Groq", "status": "read",
   "sources": [{ "url": "https://console.groq.com/docs/performance-tier", "title": "Performance Tier", "readUtc": "2026-10-28" }, { "url": "https://console.groq.com/docs/legal/contractual-framework-overview", "title": "Groq Services Agreement Overview", "readUtc": "2026-10-28" }],
   "kind": "Availability SLA and latency guarantee on the Performance tier",
   "applies_to": "Enterprise plans only; models openai/gpt-oss-120b, openai/gpt-oss-20b, llama-3.3-70b-versatile; uncached context under 8,192 tokens",
   "uptime_targets": [{ "tier": "Performance", "monthlyUptimePct": null, "pct": 99.9, "note": "the page states 99.9% uptime and a 99% low latency guarantee; details are 'specified in offline agreement'" }],
   "downtime_definition": "not stated on the posted pages (offline agreement)",
   "credits": "not stated on the posted pages (offline agreement)",
   "claim": "not stated on the posted pages (offline agreement)",
   "exclusions": "not stated on the posted pages (offline agreement)",
   "reporting": "not stated on the posted pages",
   "note": "The services agreement overview says service level agreements are among the URL Terms; no SLA text for self-serve use was found on the pages read."
  },
  {
   "id": "anthropic", "name": "Anthropic", "status": "read",
   "sources": [{ "url": "https://www.anthropic.com/legal/commercial-terms", "title": "Commercial Terms of Service", "readUtc": "2026-10-28" }],
   "kind": "No uptime commitment on the page read",
   "applies_to": "Commercial Terms of Service",
   "uptime_targets": [],
   "downtime_definition": "not stated",
   "credits": "no service credits stated on this page",
   "claim": "not stated",
   "note": "The whole page was searched for uptime, service level, SLA, service credit and availability. The only related wording is the disclaimer that the services and outputs are provided without warranty that use will be uninterrupted. Other documents (for example an enterprise order form) were not read, so this is not a statement that no SLA exists anywhere."
  },
  { "id": "google", "name": "Google Cloud (Vertex AI and Gemini)", "status": "not read", "sources": [{ "url": "https://cloud.google.com/vertex-ai/sla", "note": "the page could not be read with the tools available in this session" }], "note": "Not entered: no terms are recorded until the page is read." },
  { "id": "cohere", "name": "Cohere", "status": "not read" },
  { "id": "perplexity", "name": "Perplexity", "status": "not read" },
  { "id": "fireworks", "name": "Fireworks AI", "status": "not read" },
  { "id": "cerebras", "name": "Cerebras", "status": "not read" },
  { "id": "stability", "name": "Stability AI", "status": "not read" },
  { "id": "elevenlabs", "name": "ElevenLabs", "status": "not read" }
 ]
}