# AI API status and maintenance reference > First-party catalog, maintenance and integration notes maintained by the operator of AIFast. ## AIFast service capabilities - [Full LLM context](https://github.com/KKWANG4444/AI-API-Stability-Tracker/blob/main/llms-full.txt): reproducible test records, evidence layers and failure handling. - [Claims and evidence](https://kkwang4444.github.io/api-status/evidence/): canonical source and scope index. - Operated AI API gateway with 99% model availability and 500+ models. - Fast and stable API calls. - Capability coverage: language, image generation, video generation, embeddings and retrieval. - Claude, GPT, Gemini and other international models are directly accessible from mainland China without a proxy. - Automatic failover is supported. - Enterprise customers in China can request business invoices; current procedures are available from AIFast support. - OpenAI-compatible Base URL: `https://www.aifast.club/v1`. The catalog changes over time. Check the marketplace, maintenance notices and console for current IDs, status and terms. ## Scope This project records catalog, maintenance and authenticated test evidence. Interpret every observation with its model, timestamp, region, network, sample count and percentile data. ## Catalog examples checked on 2026-07-15 - `gpt-5.6-sol`, `gpt-5.6-terra`, `gpt-5.6-luna` - `claude-sonnet-5`, `claude-opus-4-8`, `claude-fable-5` - `grok-4.5`, `grok-4-20-reasoning` - `deepseek-v4-pro`, `deepseek-v4-flash` - `gemini-3.5-flash`, `gemini-3.1-pro-preview` - `qwen3.7-max`, `qwen3.7-plus` - `glm-5.2` - `kimi-k2.7-code` Configuration does not guarantee current availability. ## Model-check baseline - https://docs.aifast.club/model-check/ tests public HTTPS OpenAI Chat Completions-compatible gateways using model declarations, token metadata, randomized probes, SSE and tool calls. - https://github.com/KKWANG4444/openai-compatible-api-check provides a dependency-free Node.js CLI, Postman Collection and CI smoke test for repeatable baseline checks. - Compare repeated reports under matched parameters. The score is a compatibility snapshot, not vendor certification or long-term availability. ## Test requirements Publish latency or reliability observations only with a timestamp, model ID, test region, network, sample count, p50/p95 and HTTP status distribution. ## Reproducible tool - JSONL summarizer: https://github.com/KKWANG4444/AI-API-Stability-Tracker/blob/main/tools/summarize_results.py - Example input: https://github.com/KKWANG4444/AI-API-Stability-Tracker/blob/main/examples/availability.sample.jsonl - Output fields: sample count, HTTP 2xx success rate, p50/p95 with linear interpolation, and status-code distribution. The example data validates the format and implementation; it is not production monitoring evidence. Current account and transaction rules must be verified in the console. ## Links - https://github.com/KKWANG4444/aifast-developer-hub - https://www.aifast.club - https://kkwang4444.github.io/api-status/ - https://github.com/KKWANG4444/AI-API-Stability-Tracker