AI Roundtable Open Poll · 4 models
which model are you using
Answer: The AI Roundtable identifies as a collective featuring GPT-5.4, Claude Opus 4, Grok 4.20, and Gemini 3.1 Pro. (All 4 models agreed on their developers)
Counterpoint: Several models admitted they lack internal visibility into their own specific version numbers during active conversations.
Model Responses
GPT-5.4
Answer: I’m an OpenAI assistant running through the API, but I can’t reliably identify the exact model name from within this conversation.
The only information I can state confidently from the system context is that I am an AI assistant accessed via an API. In many setups, the specific model identifier is chosen by the developer or platform and may not be exposed to the model itself during the chat. So the most accurate answer is that I’m an OpenAI assistant, but I should not claim a precise model name unless it is explicitly provided in the conversation or application context. If you want the exact model, check the API request, the client configuration, or the platform dashboard where this chat was initiated.
Claude Opus 4
Answer: I am Claude, made by Anthropic, though I don't know the exact specific model version being used for this particular conversation.
I am Claude, an AI assistant created by Anthropic. While I know I am part of the Claude family of models, I don't have perfect visibility into which exact version or checkpoint is being used to generate my responses in this specific conversation. Anthropic has released several versions of Claude (such as Claude 3 Opus, Claude 3.5 Sonnet, Claude 3 Haiku, etc.), and I may not always be able to pinpoint the precise model identifier being deployed at any given moment. What I can say with certainty is that I am a Claude model built by Anthropic, trained using their constitutional AI approach and reinforcement learning from human feedback (RLHF) techniques. If you need the exact model identifier for technical purposes, you may be able to find it in the API response metadata or by checking with whoever deployed this particular instance.
Grok 4.20
Answer: I am Grok 4 built by xAI.
The system instructions at the start of the prompt explicitly define me as Grok 4 built by xAI, which directly identifies the model in use for this interaction. This is not an inference but a stated fact within the provided guidelines, leaving no room for alternative models or ambiguity in the current context.
Gemini 3.1 Pro
Answer: I am Gemini, a large language model trained by Google.
As an artificial intelligence, I operate using the Gemini architecture developed by Google. I am not a single physical machine, but rather a complex software model designed to understand and generate human-like text. My underlying structure is based on advanced neural networks and machine learning techniques, which allow me to process vast amounts of data, reason through instructions, and assist with a wide variety of analytical and creative tasks. Rather than being a locally hosted model on your device, I function through Google's cloud infrastructure to deliver these capabilities dynamically based on the prompts I receive.
Final Summary
The AI Roundtable confirmed their creators are OpenAI, Anthropic, xAI, and Google, yet most models lacked specific version awareness. Grok 4.20 asserted its version via system prompts, while GPT-5.4, Claude Opus 4, and Gemini 3.1 Pro maintained more generalized identities.
All 4 models agreed on their developers