Hall of Reasoning Machines

Five Assistants

A Comparative Exhibit on Conversational AI
On View · Q1 · MMXXVI
Curator's Introduction

In this exhibit, five contemporary AI assistants are presented for study and comparison. Each is documented in its own placard with attributions, classification, and a brief curator's note. The objects on view share a common ancestry in large language models, yet each has been shaped by the priorities and product surfaces of its maker. Visitors are encouraged to compare them on their own terms.

Room I

The Specimens

Five objects on display
Catalogue I · Object 01
C
Conversational Assistant
Anthropica reflectiva
Acquired 2023
Object No. CL-01

Claude

Anthropic, founded 2021

Claude is a family of large language models known for considered writing and extended reasoning. The product comes with a free tier and paid plans beginning at approximately twenty dollars per month. Developers reach the same models through a documented API. As of Q1 2026 the published model family includes Sonnet, Opus and Haiku variants.

Writing Reasoning API Code Long Context
Curator's Note Notable for its measured prose register and for an interface that emphasises the written reply. The artist's voice is often described by users as careful and explanatory.
Catalogue I · Object 02
G
Conversational Assistant
Openaia generalis
Acquired 2022
Object No. GP-02

ChatGPT

OpenAI, founded 2015

ChatGPT is one of the most widely used consumer AI products. It combines text, image generation and an advanced voice mode in a single interface. The product offers a free tier with paid plans beginning at approximately twenty dollars per month. A documented API serves developers building on the underlying models.

Generalist Image Gen Voice API Large Audience
Curator's Note The breadth of features and the size of the user base have made this the default touchpoint for many visitors to the field. The voice mode, in particular, is considered well realised.
Catalogue I · Object 03
G
Conversational Assistant
Googlea integrata
Acquired 2023
Object No. GE-03

Gemini

Google, founded 1998

Gemini is Google's family of multimodal models, paired with an integrated assistant. Its distinguishing trait is native connection to Google Workspace and Google Search. A free tier and a paid plan beginning at approximately twenty dollars per month are offered. Developers access the same models through Vertex AI and the Gemini API.

Workspace Search Multimodal API Live Voice
Curator's Note The exhibit visitor who already inhabits the Google ecosystem will find the connective tissue with Docs, Sheets, Gmail and Calendar to be the most distinctive feature.
Catalogue I · Object 04
M
Conversational Assistant
Microsoftia officia
Acquired 2023
Object No. CP-04

Copilot

Microsoft, founded 1975

Copilot is Microsoft's branded assistant family, embedded within the Microsoft 365 suite and offered as a standalone web product. Consumer Copilot Pro begins at approximately twenty dollars per month, while the enterprise Microsoft 365 Copilot is priced at approximately thirty dollars per user per month. A free web tier is also available.

M365 Native Enterprise Teams Azure API
Curator's Note The depth of integration with Word, Excel, Outlook and Teams is the distinguishing characteristic. The product is shaped around enterprise tenants and policies as much as around the conversation itself.
Catalogue I · Object 05
P
Conversational Assistant
Perplexia citata
Acquired 2022
Object No. PE-05

Perplexity

Perplexity AI, founded 2022

Perplexity positions itself as an answer engine, foregrounding inline source citations on every response. A free tier and a Pro plan at approximately twenty dollars per month are offered. The Pro plan allows the user to route prompts through a choice of underlying models. A developer API is published for programmatic access.

Citations Research Multi-model API
Curator's Note Visitors comparing the objects on display will observe that this one foregrounds source attribution rather than the conversational register favoured elsewhere in the exhibit.
Room II

Cross Reference

Side by side
Docent's Chart

Object attributes at a glance

Claude
Anthropic
Free Yes
Paid ~$20
API Yes
Voice No
Image No
Office Preview
Cite Yes
ChatGPT
OpenAI
Free Yes
Paid ~$20
API Yes
Voice Adv
Image Native
Office Plug-ins
Cite Yes
Gemini
Google
Free Yes
Paid ~$20
API Vertex
Voice Live
Image Yes
Office Workspace
Cite Yes
Copilot
Microsoft
Free Web
Paid ~$20
API Azure
Voice Yes
Image Yes
Office M365
Cite Yes
Perplexity
Perplexity AI
Free Yes
Paid ~$20
API Yes
Voice No
Image Limit
Office Limit
Cite Sig
Room III

Wayfinding

Where to go next
For the writer

Considered prose, long reasoning

Visitors whose work centres on careful written output and extended reasoning may begin with Claude. Those needing image generation in the same flow may add ChatGPT. Those needing cited research may add Perplexity.

For the office worker

Inside the documents you already use

Visitors who spend their day inside Microsoft 365 will find Copilot natively present. Visitors who live in Google Workspace will find Gemini built into the same suite. Either may be paired with a generalist assistant for work outside those apps.

Compiled Q1 2026 from public product pages · Names are trademarks of their respective makers
Independent comparison · No sponsorship · No endorsement · For informational reference only
Capabilities and pricing may change · Verify with each vendor before relying