AI products guide

The most widely used AI chatbot: conversation, writing, image generation, voice, file analysis and agents.

• Conversation, writing, translation and explanation
• Searching the web for current information
• Creating and editing images inside the chat

Pricing: Free version; paid plans listed on OpenAI's site

An assistant known for writing, long documents and coding, which can also carry out tasks on your computer.

• Writing, editing and analysing long documents
• Web search, running code and creating files (documents, slides, spreadsheets)
• Memory across conversations

Pricing: Free · Pro $20/month ($17/month billed annually) · Max from $100/month

Microsoft's assistant across Windows, Edge, Bing and Microsoft 365. It began as Bing Chat in February 2023.

• Chat and web search through Bing
• Built into Windows and the Edge browser
• Microsoft 365 Copilot drafts documents, analyses spreadsheets and summarises meetings and email

A free Chinese assistant built on DeepSeek's own open-weight models, known for strong reasoning at low cost.

• Conversation, writing and coding
• Step-by-step reasoning for maths and logic
• 1-million-token context with the V4 models

Pricing: Free app; the API is paid per use

ByteDance's AI assistant, China's most popular chatbot, with 330 million users by May 2026.

• Text, image and voice generation
• AI-powered search
• In-car voice assistant in Tesla cars sold in mainland China

Google's assistant, built into Android, Gmail, Docs and Search, with image and video generation.

• Conversation and writing
• Deep Research: multi-step research reports
• Gemini Live: real-time voice conversation

Pricing: Free · Google AI Plus $4.99/month · Google AI Pro $19.99/month · Ultra (US prices)

The chatbot from SpaceXAI (formerly xAI), integrated with the X social network.

• Conversation with access to posts on X
• Web search and a reasoning ("Think") mode
• Image generation (Grok Imagine)

Moonshot AI's assistant and open-weight models. Kimi K3 is the largest open-weight model, at 2.8 trillion parameters.

• Conversation and reading long documents
• Image and video input
• "Swarm" mode uses hundreds of sub-agents for research

The assistant from France's Mistral AI, Europe's most valuable AI company. Launched as Le Chat in 2024 and renamed Mistral Vibe in 2026.

• Conversation, writing and coding
• Several Mistral models are released as open weights

Alibaba's assistant and model family, most of which is released as open weights.

• Conversation, writing and coding
• Real-time voice chat with the Omni models (text, image, video and audio input)
• Generating text, pictures or video with Qwen3-Max-Thinking

Upload your own documents and ask questions about them, or turn them into a podcast-style audio discussion.

• Summaries, explanations and answers based on your uploaded material
• Audio Overviews: a podcast-style conversation about your sources

An "answer engine" that searches the web and replies in full sentences with citations.

• Answers with numbered sources you can open
• Searching inside your uploaded Excel, Word and PDF files
• Perplexity Assistant (January 2025) can take actions in other apps, such as booking a ride

Pricing: Free version; paid plans listed on Perplexity's site

Adobe's creative AI: images, video, speech, soundtracks and sound effects, also built into Photoshop and Premiere Pro.

• Text to image and text to video
• Image to video
• Generating speech, soundtracks and sound effects

Pricing: Subscription plans with monthly credits

An independent lab's image generator known for an artistic, polished look. Open to the public since July 2022.

• Generating images from text descriptions (prompts)
• Many artistic styles

Google's image model inside the Gemini app, for creating and editing pictures with plain instructions.

• Change lighting, camera angle and focus of a photo
• Apply the style of a reference image
• Combine two photos into one

Pricing: Included in Gemini; paid plans have higher limits

ByteDance's image and video creation platform, from the makers of CapCut. Its Seedance video models rank near the top of LMArena's video leaderboard.

• Video from images: product showcases, social clips, character animation
• Posters, logos, app icons and stickers
• Restoring old photos, swapping backgrounds, restyling portraits

The video generation service from Shanghai's MiniMax, one of China's six "AI tigers", listed in Hong Kong in January 2026.

• Video from text or images
• Text and music generation on the Hailuo platform

Kuaishou's video generator, first released in June 2024, now with an agent that helps write scripts and storyboards.

• Video from text or images
• Agent that brainstorms, writes scripts and generates storyboards and posters
• Multiple rounds of editing within one project

A New York company making video and image generation tools for creators and studios.

• Generating and editing video and images
• Used in films such as Everything Everywhere All at Once and on The Late Show with Stephen Colbert

Google's video models, used through the Gemini app and the Flow filmmaking studio. Veo 3 (May 2025) generates sound along with video.

• Video from text or images, with dialogue and sound effects
• Flow: build cinematic scenes and stories from multiple clips

Pricing: Included in Google AI paid plans, with credits

Natural-sounding AI voices for narration, dubbing and audiobooks. Founded in 2022 and based in London.

• Text to speech in many voices and languages
• Dubbing videos into other languages
• Reader app: listen to articles, PDFs and e-books

Pricing: Free plan with 10,000 credits a month; paid plans on the official site

Creates full songs with vocals from a text, audio, image or video prompt.

• Complete songs with vocals and instruments
• Suno Studio for editing, with MIDI and effects (top plan)
• v6 models (September 2026) developed with Warner Music Group, BMG and Believe

Pricing: Free plan; paid plans include commercial use rights

Google's platform where AI agents write, run and test code, launched with Gemini 3 in November 2025.

• Agents that write, run and test code with varying autonomy
• Antigravity 2.0 (May 2026): a desktop app for managing agents, plus a CLI and SDK
• Choice of Gemini, Anthropic and OpenAI models

Pricing: Free within usage limits

An agent that works on whole software projects from the terminal or an editor: reading, changing, running and testing code.

• Reading and changing code across a whole project
• Running commands and tests to check its work
• Working on large, long-running tasks

Pricing: Included with Claude Pro and Max

OpenAI's coding agent, with more than 2 million weekly users by March 2026.

• Writing code and fixing bugs
• Running commands and editing files
• Codex Security: finding and fixing vulnerabilities

An AI code editor and coding agent. Its maker passed $3 billion in annual revenue in early 2026 and became part of SpaceXAI in August.

• Agent searches the codebase, edits files, runs terminal commands and completes multi-step tasks
• Agents for web, mobile, command line and cloud

Pricing: Hobby free · Individual $20/month · Teams $40/user/month

The first widely used AI coding assistant, announced in June 2021.

• Code completion as you type
• Lets you choose the underlying model

Pricing: Subscription for individuals and businesses

A Swedish "vibe coding" platform: describe an app or website and it builds it for you, no programming needed.

• Building web apps and sites from plain-language prompts

An autonomous agent that researches, browses and produces finished work such as reports. Developed by a company founded in China and now based in Singapore.

• Multi-step research and web tasks
• Producing finished deliverables such as reports and analyses

AI built into Notion, the notes and project workspace: agents, meeting notes and search across connected apps.

• Notion Agent completes multi-step tasks using your workspace, connected apps and the web
• AI Meeting Notes without a meeting bot
• Enterprise Search across Slack, GitHub and other apps

Pricing: Free trial; agents cost $10 per 1,000 credits after the trial