Independent reporting and reviews on the consumer technology people actually use.
AI TOOLS VERTICAL
Reporting on consumer-facing generative AI: chatbots, AI photo and image tools, AI productivity assistants, on-device AI features, and the policy questions around all of it.
SAN FRANCISCO · JUN 27
A cluster of price cuts and raised usage limits across the AI pair-programming category in June has analysts describing the market as a full price war, with vendors trading cheaper monthly plans and higher request caps to hold onto developers.
SAN FRANCISCO · MAY 15
The AI-first code editor on Thursday released Background Agents 1.0, a long-running task system that can plan, execute, and review changes across multiple repositories without an active editor session. The company said the feature is its most direct answer yet to Claude Code's terminal-native workflow.
SAN FRANCISCO · MAY 8
OpenAI on Thursday cut input and output token prices on its o3 reasoning-model family by roughly 40% and doubled rate limits across paid API tiers, the company said, citing inference-stack improvements as the primary driver.
SAN FRANCISCO · MAY 7
Web search is now switched on by default for Claude.ai Pro and Team accounts. Answers that use it cite their sources automatically, and publishers get an opt-in registry modeled on the one Anthropic built for its ClaudeBot crawler, the company told Consumer Tech Wire.
SAN FRANCISCO · APR 9
The AI-search company on Thursday introduced a higher-tier offering that lets users restrict queries to a curated allowlist of publishers and academic sources — a response, executives said, to enterprise demand for provenance controls.
PALO ALTO · MAR 18
The new model brings native multimodal reasoning across text, image, and audio and a substantially extended context window, OpenAI said in a blog post Wednesday — the company's first major model upgrade since GPT-4o launched in 2024.
MOUNTAIN VIEW · MAR 3
Google on Tuesday released Gemini 3 across its Pro and Ultra tiers, restructuring the Gemini lineup around a single trained model family and a 2-million-token context window for paid customers.
SAN FRANCISCO · FEB 11
One million developers now pay for Cursor, its maker Anysphere said Tuesday, alongside a revenue run rate north of $480 million: a measure of how quickly an AI-first editor has pulled users away from the Visual Studio Code setup it was forked from.
SAN FRANCISCO · JAN 21
Opus 4 arrived Tuesday as Anthropic's biggest model since Opus 3, and it reasons before it answers unless you tell it not to: extended thinking now ships switched on for paying Claude.ai users and for Opus-tier API traffic.
NEW YORK · DEC 17
OpenAI on Wednesday expanded the rollout of its SearchGPT crawler, OAI-SearchBot, to all eligible web publishers — a move that several news organizations said they would block by default, arguing the company has not yet established acceptable revenue-share terms.
REVIEW · BOSTON · JAN 15
Consumer Tech Wire's 2026 ranking of consumer AI chatbots, scored on reasoning quality, response accuracy, multimodal capability, context handling, and price. Six leading assistants tested across 220 evaluation prompts.
ANALYSIS · SAN FRANCISCO · APR 2
OpenAI's GPT-5 release in February pulled the headline-benchmark crown back from Anthropic. The benchmarks tell one story; six weeks of side-by-side daily use tells a different one.
GUIDE · SAN FRANCISCO · MAR 22
A practical guide to picking a primary AI chatbot for daily knowledge work in 2026, organized around six work profiles and the model that actually fits each one.