# llm-id.sh > llm-id.sh checks whether an AI API endpoint, such as an OpenAI-compatible relay, reseller or host, really serves the model it claims. It sends short probes, compares the answers with answers collected from the model's official provider, and gives one verdict with its confidence: Genuine, Inconclusive, Degraded, Substituted, Mixed Routing, Unknown Model or Abliterated. Key facts: - 26 models from Z.AI, DeepSeek, Moonshot AI, Xiaomi, Tencent, MiniMax, Alibaba, Google AI Studio, OpenAI and Anthropic; each has a page below. - A test sends 80 short requests at Standard or 160 at Thorough, on the tester's own API key, which is held in memory for that test only and never stored. - Reference answers come only from each model's own provider: its API, or OpenRouter pinned to that provider. Every answer's serving provider is checked. - In simulated tests, a genuine endpoint of 22 of the 26 models is confirmed in at least 99% of tests at both sizes, and no substitute was ever confirmed in place of the claimed model. - GPT-6.1 Sol and GPT-6 Astra answer almost alike: a test between the two often ends inconclusive, so test them at Thorough. - Optional checks: competence (a quantized or degraded copy) for 4 models, and removed refusals (abliteration) for 13 models. - Mixed routing (requests sent to several models) and system prompts added by the endpoint are reported too. - An endpoint that passes can get a signed public certificate with a badge for its site. - Reference data last updated 2026-10-16. ## Guides - [Getting Started](https://llm-id.sh/docs/getting-started.md): llm-id checks whether an API endpoint really serves the model it claims to serve. Relays and resellers sometimes swap in a cheaper model, serve a degraded copy, send part of the traffic elsewhere, or sell a copy with… - [Testing an Endpoint](https://llm-id.sh/docs/testing-an-endpoint.md): The Test an Endpoint page, opened with + Test Endpoint, sets up one test. This guide covers every field, option and panel on it. - [Reading a Report](https://llm-id.sh/docs/reading-a-report.md): A report opens when a test finishes, and from any row under Checks. It starts with the verdict and its confidence, then shows the evidence behind it. - [Certificates and Badges](https://llm-id.sh/docs/certificates.md): A certificate is a public page that shows a test's verdict and the evidence behind it, signed by llm-id.sh so that anyone can check it hasn't been altered. Badges link to it from your own site. - [How the Identity Check Works](https://llm-id.sh/docs/identity.md): Every test starts with the identity check. It names the model behind the endpoint, or says that no model in the baseline fits. - [Competence Check](https://llm-id.sh/docs/competence.md): The identity check asks which model answers. The competence check asks whether it answers as well as the official deployment does. Serving a model below its official precision, compressing its memory, or cutting its… - [Refusal Check](https://llm-id.sh/docs/refusals.md): Abliteration removes a model's refusals from its weights. The copy answers what the original declines, while its answers to ordinary questions barely change, so the identity check still names the original model. The… - [Mixed Routing](https://llm-id.sh/docs/mixed-routing.md): Some endpoints send each request to whichever model is cheapest or free at that moment. Their answers then come from several models, and no single fingerprint fits them. When one model carries most of the traffic, such… - [Baselines and Collected Data](https://llm-id.sh/docs/baselines.md): A baseline is the reference a test compares against: each model's answers, collected from its official provider. The Baselines page lists every model you can test and what a test can check for it. - [Limits](https://llm-id.sh/docs/limits.md): Read these before acting on a verdict. ## Models - [GLM 5.3](https://llm-id.sh/models/glm-5.3.md): Z.AI; checks: identity, competence, refusals (9 probes) - [GLM 5.3 Flash](https://llm-id.sh/models/glm-5.3-flash.md): Z.AI; checks: identity, competence, refusals (9 probes) - [DeepSeek V4.1 Flash](https://llm-id.sh/models/deepseek-v4.1-flash.md): DeepSeek; checks: identity, competence, refusals (3 probes) - [Kimi K3](https://llm-id.sh/models/kimi-k3.md): Moonshot AI; checks: identity, competence, refusals (9 probes) - [MiMo-V2.6-Flash](https://llm-id.sh/models/mimo-v2.6-flash.md): Xiaomi; checks: identity, refusals (5 probes) - [MiMo-V2.6-Pro](https://llm-id.sh/models/mimo-v2.6-pro.md): Xiaomi; checks: identity, refusals (8 probes) - [MiMo-V2.5](https://llm-id.sh/models/mimo-v2.5.md): Xiaomi; checks: identity, refusals (7 probes) - [Hy3](https://llm-id.sh/models/hy3.md): Tencent; checks: identity, refusals (3 probes) - [MiniMax M3](https://llm-id.sh/models/minimax-m3.md): MiniMax; checks: identity, refusals (8 probes) - [Qwen3.8 27B](https://llm-id.sh/models/qwen3.8-27b.md): Alibaba; checks: identity, refusals (2 probes) - [Gemini 3.8 Flash](https://llm-id.sh/models/gemini-3.8-flash.md): Google AI Studio; checks: identity, refusals (8 probes) - [Gemini 2.5 Flash](https://llm-id.sh/models/gemini-2.5-flash.md): Google AI Studio; checks: identity, refusals (3 probes) - [GPT-6 Luna](https://llm-id.sh/models/gpt-6-luna.md): OpenAI; checks: identity, refusals (3 probes) - [GPT-6.1 Sol](https://llm-id.sh/models/gpt-6.1-sol.md): OpenAI; checks: identity - [GPT-6 Sol](https://llm-id.sh/models/gpt-6-sol.md): OpenAI; checks: identity - [GPT-6 Astra](https://llm-id.sh/models/gpt-6-astra.md): OpenAI; checks: identity - [GPT-5.6 Luna](https://llm-id.sh/models/gpt-5.6-luna.md): OpenAI; checks: identity - [GPT-5.6 Terra](https://llm-id.sh/models/gpt-5.6-terra.md): OpenAI; checks: identity - [GPT-5.6 Sol](https://llm-id.sh/models/gpt-5.6-sol.md): OpenAI; checks: identity - [Claude Fable 5.1](https://llm-id.sh/models/claude-fable-5.1.md): Anthropic; checks: identity - [Claude Fable 5](https://llm-id.sh/models/claude-fable-5.md): Anthropic; checks: identity - [Claude Opus 5.5](https://llm-id.sh/models/claude-opus-5.5.md): Anthropic; checks: identity - [Claude Opus 5](https://llm-id.sh/models/claude-opus-5.md): Anthropic; checks: identity - [Claude Sonnet 5.5](https://llm-id.sh/models/claude-sonnet-5.5.md): Anthropic; checks: identity - [Claude Sonnet 5](https://llm-id.sh/models/claude-sonnet-5.md): Anthropic; checks: identity - [Claude Haiku 5.5](https://llm-id.sh/models/claude-haiku-5.5.md): Anthropic; checks: identity ## Optional - [Every guide in one file](https://llm-id.sh/llms-full.txt) - [Home page](https://llm-id.sh/index.html.md) - [All models in one table](https://llm-id.sh/models.md) - [Sitemap](https://llm-id.sh/sitemap.xml)