EPISODE · THURSDAY, MARCH 5, 2026

Gemini 3.1 Flash Lite Beats GPT-5 Mini, Claude Code Goes Voice, and Bing Makes GEO Official

Google launches Gemini 3.1 Flash Lite — 2.5x faster than previous version, outperforming GPT-5 mini and Claude 4.5 Haiku at $0.25/1M tokens. Claude hits $2.5B run-rate and overtakes ChatGPT in App Store. Apple's MacBook Neo launches at $599 with A18 Pro chip. Bing officially adds GEO to webmaster guidelines with opt-out mechanism. DeepSeek V4 drops with 1 trillion parameters and 1M context window.

Ready to Play

0:00 / 8:10

In This Episode

  • • Gemini 3.1 Flash Lite is 2.5x faster than Gemini 2.5 Flash at just $0.25/1M input tokens
  • • Gemini 3.1 Flash Lite outperforms GPT-5 mini and Claude 4.5 Haiku on 6 of 11 benchmarks
  • • Claude hits $2.5B annual run-rate with weekly users doubled since January
  • • Claude overtakes ChatGPT in App Store rankings
  • • Claude Code Voice Mode rolling out to 5% of users — speak coding commands directly
  • • MacBook Neo launches at $599 with A18 Pro chip — 50% faster, 16-hour battery
  • • Bing officially adds GEO to webmaster guidelines — first major search engine to do so
  • • DeepSeek V4: 1 trillion parameters, 1M context window, sparse MoE architecture

Show Notes

  • Gemini 3.1 Flash Lite benchmarks and pricing
  • Claude $2.5B run-rate and App Store overtake
  • Claude Code Voice Mode rollout
  • GPT-5.3 Instant update
  • MacBook Neo at $599 with A18 Pro chip
  • Bing adds GEO to webmaster guidelines
  • GEO citation lift data — statistics, experts, sources
  • DeepSeek V4 — 1T parameters, 1M context
  • New tools: Base44, Wispr Flow, Bolt.new, nexos.ai

Transcript

[00:00] Introduction

Welcome to the Daily AI Digest for Thursday, March 5th, 2026.

Today: Gemini 3.1 Flash Lite outperforms GPT-5 mini at a fraction of the price. Claude hits $2.5B run-rate and overtakes ChatGPT in the App Store. Apple's MacBook Neo launches at $599. Bing officially adds GEO to its webmaster guidelines. And DeepSeek drops a one-trillion parameter model.

[00:45] Gemini 3.1 Flash Lite

Google launched Gemini 3.1 Flash Lite — 2.5x faster than Gemini 2.5 Flash, priced at $0.25/1M input tokens and $1.50/1M output tokens.

It outperforms GPT-5 mini and Claude 4.5 Haiku on 6 of 11 standard benchmarks — significant cost reduction without sacrificing capability.

Google also launched a Workspace CLI for Drive, Gmail, Calendar, Sheets, Docs, Chat, and Admin — built with AI agents in mind.

[01:45] Claude $2.5B Run-Rate & Voice Mode

Anthropic's Claude is running at a $2.5 billion annual revenue run-rate. Weekly users doubled since January. Claude has overtaken ChatGPT in App Store rankings.

Claude Code Voice Mode is now rolling out to 5% of users — speak commands like 'refactor the authentication middleware' and Claude executes them directly.

OpenAI updated ChatGPT to GPT-5.3 Instant — fewer hallucinations, fewer refusals, better web search. Full GPT-5.3 'Garlic' API launch expected soon.

[03:00] MacBook Neo at $599

Apple launched the MacBook Neo at $599 — the first Mac with the A18 Pro chip (same as iPhone 16 Pro). 50% faster than Intel Ultra 5, 16-hour battery life.

At $599, it undercuts the MacBook Air by $400 and brings Apple Silicon's Neural Engine to a much broader consumer segment.

Google Play Store fees also dropped to 20% (15% for new installs), with third-party app stores now able to apply for streamlined Android installation.

[03:45] Bing Makes GEO Official

Bing added Generative Engine Optimization to its official webmaster guidelines — the first major search engine to formally acknowledge GEO as a distinct practice.

Site owners can now opt out of AI training through Bing's webmaster tools.

GEO citation lift data: adding statistics lifts AI citations by up to 33.9%, quoting named experts adds up to 32%, citing sources adds up to 30.3%.

Context: AI Overviews appear on 88–99% of informational searches. CTR drops up to 61% when AI Overviews appear. Only 374 clicks per 1,000 US searches reach the open web. 85% of AI Overview citations come from sources with 3+ E-E-A-T signals.

[05:30] DeepSeek V4

DeepSeek V4: 1 trillion parameters, sparse Mixture-of-Experts architecture, 1 million token context window.

The sparse MoE design keeps inference costs manageable despite the enormous parameter count — decoupling parameter count from inference cost.

[06:00] New Tools

Base44: describe any app in plain English, it builds UI, database, auth, and hosting automatically.

Wispr Flow: voice dictation with 89% of messages sent with zero edits, 4x faster than typing.

Bolt.new: added built-in AI image generation with transparent backgrounds directly in the coding environment.

nexos.ai: no-code AI agents connecting to Google Drive, Gmail, and Slack with deep research capabilities.

OpenAI Codex is now available as a native Windows app via the Microsoft Store.

[06:45] What This Means For You

Test Gemini 3.1 Flash Lite against your current model — at $0.25/1M input tokens, cost savings on high-volume workloads could be substantial.

Audit your most important content pages for E-E-A-T signals: statistics, named expert quotes, and cited sources. 85% of AI citations go to high E-E-A-T content.

If you use Claude Code, watch for the voice mode rollout — currently at 5% of users and expanding.

[07:30] Closing

That's your Daily AI Digest for Thursday, March 5th, 2026. Gemini 3.1 Flash Lite reshapes the cost curve for AI APIs. Claude overtakes ChatGPT in the App Store. Bing formalizes GEO. And DeepSeek pushes the parameter frontier to one trillion.

Stay informed, stay ahead — and we'll see you tomorrow.

Related Episodes

More episodes you might enjoy based on similar topics and categories.

AI Models & Research

Wednesday, March 25, 2026

Siri Becomes an Agent, OpenAI Kills Sora, and Why 97% of AI Content Fails After 3 Months

Apple is testing a standalone Siri app for iOS 27 and macOS 27 that transforms Siri from a voice assistant into a system-wide AI agent — unveiling expected at WWDC June 8. OpenAI has shut down Sora just months after launch; the real reason is compute economics: $13B revenue vs $100B projected spend. A 16-month experiment shows 97% of AI-generated pages fell out of the top 100 after 3 months without authority signals. Plus: Claude Code auto mode, MiniMax M2.7 engineering agent, the SaaS revenue durability crisis, and the end of free trials.

AnthropicOpenAIGoogleAgentsEnterprise

9:10

AI Models & Research

Saturday, March 7, 2026

GPT-5.4 Goes Agentic, Codex Finds 800 Bugs, and the 90.9% GEO Benchmark

OpenAI releases GPT-5.4 with 1M token context and 75% desktop automation. Codex Security finds 800 critical bugs in OpenSSH and Chromium. Samsung AI glasses enter wearables race. Netflix acquires Ben Affleck's AI filmmaking startup. GenOptima sets 90.9% AI recommendation rate as the new GEO benchmark.

OpenAISEOLLMsSecurityAgents

5:46

AI Models & Research

Wednesday, March 18, 2026

GPT-5.4 Mini Costs Less Than a Stamp, Google Now Reads Your Gmail, and Why Faster Code Is Making Teams Slower

OpenAI releases GPT-5.4 mini ($0.75/M tokens) and nano ($0.20/M tokens) — production models designed for high-volume workflows. Google expands Personal Intelligence across Search, Gemini, and Chrome, pulling from Gmail and Photos to personalise results. AI overviews cut click-through rates by 58%, but AI-referred visitors convert at higher rates. Stripe's coding agents push 1,300 PRs/week using blueprint hybrid workflows. And the Atrophy Paradox: heavy AI use reduces cognitive engagement and makes output converge.

OpenAIGoogleSEOAgentsMarketing

7:10

AI Models & Research

Friday, March 13, 2026

GPT-5.4 Cites Brand Sites 56% of the Time, Google Maps Gets AI, and the Responses API Changes Everything

GPT-5.4 now cites brand sites 56% of the time — up 7x from GPT-5.3. Pricing pages cited 34x more. Google launches Ask Maps with Gemini. OpenAI Responses API gives agents full computer access. Meta's Avocado model delayed to May. Anthropic study: AI impacts knowledge workers most.

OpenAIGoogleSEOGEOAgents

5:48