deployedbyai

The Deploy Log | Month by month | Free, no signup

What shipped in AI in October 2025

49 rows are listed for October 2025 in The Deploy Log: 10 leads, 20 in Models and APIs, 9 in Tools and products, 3 in Robotics, hardware and chips, 5 in Breakthroughs and 2 in Funding and acquisitions, from 9 publishers. They are listed from to . They include Company knowledge from OpenAI, Mistral AI Studio from Mistral and 47 others. Every item sourced to the publisher's own page.

In figures

Companies with the most rows
CompanyRows
OpenAI15
Vercel9
Hugging Face8
Anthropic6
Google Research5
Waymo3
Google DeepMind1
Mistral1
Y Combinator1
FigureValue
Rows49
Rows the month before21
Change against the month before+28
Companies9
Companies seen for the first time4: Google DeepMind, Google Research, Mistral and Y Combinator
Models seen for the first time2: Claude Haiku 4.5 and Claude Sonnet 4.5
Rows with no shipped date, placed by their edition or the day they were added30
Rows by week, October 2025Week of September 29, 2025: 8 rows; Week of October 6, 2025: 6 rows; Week of October 13, 2025: 13 rows; Week of October 20, 2025: 18 rows; Week of October 27, 2025: 4 rows.Sep 29Oct 6Oct 13Oct 20Oct 278613184
Each bar counts the rows of this month listed in that week. A week that runs into another month counts only this month's rows. Data as of .

Leads

Company knowledge

Company knowledge for ChatGPT Business, Enterprise, and Edu brings context from connected apps like Slack, SharePoint, Google Drive, and GitHub into answers with citations, powered by a version of GPT-5 trained to look across multiple sources.

OpenAI | In Edition 6,

SOURCE OpenAI, Work smarter with your company knowledge in ChatGPT

Mistral AI Studio

Mistral AI Studio packages Observability, Agent Runtime, and AI Registry into one platform for building, evaluating, and running AI in production, with hybrid, dedicated, and self-hosted deployments.

Mistral | In Edition 6,

SOURCE Mistral, Introducing Mistral AI Studio

Vercel Agent

Vercel Agent is available in Public Beta with Code Review and Investigations, running simulated builds inside a Vercel Sandbox to verify pull request recommendations and analyzing telemetry to explain production anomalies.

Vercel | In Edition 6,

SOURCE Vercel, Introducing Vercel Agent

Claude Haiku 4.5

Claude Haiku 4.5 is available today to all users at $1 per million input tokens and $5 per million output tokens, with coding performance similar to Claude Sonnet 4 at one third the cost and more than twice the speed.

Anthropic | In Edition 5,

SOURCE Anthropic, Introducing Claude Haiku 4.5

DoorDash delivery by Waymo

Waymo and DoorDash are partnering to match DoorDash customers with fully autonomous Waymo vehicles for deliveries from DashMart across a 315 square mile service area in Phoenix, with more local merchants planned over time.

Waymo | In Edition 5,

SOURCE Waymo, Your DoorDash Order, Delivered by Waymo

AgentKit

AgentKit launched with Agent Builder in beta, ChatKit generally available, and four new Evals capabilities: datasets, trace grading, automated prompt optimization, and third-party model support. Ramp cut iteration cycles by 70%.

OpenAI | In Edition 4,

SOURCE OpenAI, Introducing AgentKit

Apps in ChatGPT

Apps in ChatGPT launched with the Apps SDK preview, an open standard built on the Model Context Protocol. Booking.com, Canva, Coursera, Figma, Expedia, Spotify, and Zillow are live today for logged-in users outside the EEA, Switzerland, and the UK.

OpenAI | In Edition 4,

SOURCE OpenAI, Introducing apps in ChatGPT and the new Apps SDK

Claude Sonnet 4.5

Claude Sonnet 4.5 is available everywhere today at $3/$15 per million tokens, the same price as Claude Sonnet 4. It leads OSWorld at 61.4% and maintains focus for more than 30 hours on complex tasks.

Anthropic | In Edition 3,

SOURCE Anthropic, Introducing Claude Sonnet 4.5

Instant Checkout

U.S. ChatGPT Plus, Pro, and Free users can now buy directly from U.S. Etsy sellers in chat, with over a million Shopify merchants like Glossier, SKIMS, Spanx and Vuori coming soon. The Agentic Commerce Protocol is open sourced today.

OpenAI | In Edition 3,

SOURCE OpenAI, Buy it in ChatGPT: Instant Checkout and the Agentic Commerce Protocol

Sora 2

Sora 2 is a video and audio generation model with synchronized dialogue and sound effects, more accurate physics, and a characters feature that inserts a recorded person into generated scenes. The Sora iOS app is available to download now with initial rollout in the U.S. and Canada.

OpenAI | In Edition 3,

SOURCE OpenAI, Sora 2 is here

Models and APIs

gpt-oss-safeguard

OpenAI released gpt-oss-safeguard-120b and 20b, open-weight reasoning models for safety classification under Apache 2.0, downloadable from Hugging Face.

OpenAI | Models and APIs | Published

SOURCE OpenAI, Introducing gpt-oss-safeguard

ChatGPT sensitive conversations

OpenAI updated ChatGPT's default model to reduce undesired responses in sensitive conversations by 65-80%, working with over 170 mental health experts.

OpenAI | Models and APIs | Published

SOURCE OpenAI, Strengthening ChatGPT’s responses in sensitive conversations

Claude for Life Sciences

Claude Sonnet 4.5 scores 0.83 on Protocol QA against a human baseline of 0.79, and Anthropic is adding connectors to Benchling, BioRender, PubMed, Scholar Gateway, Synapse.org, and 10x Genomics plus a single-cell-rna-qc Agent Skill.

Anthropic | Models and APIs | In Edition 6,

SOURCE Anthropic, Claude for Life Sciences

LeRobot v0.4.0

LeRobot v0.4.0 introduces Datasets v3.0 with chunked episodes and streaming, integrates PI0, PI0.5, and GR00T N1.5 policies, adds LIBERO and Meta-World simulation support, and launches a plugin system for third-party hardware.

Hugging Face | Models and APIs | In Edition 6,

SOURCE Hugging Face, LeRobot v0.4.0: Supercharging OSS Robot Learning

MedGemma 27B Multimodal

MedGemma 27B Multimodal adds image and text inputs, scoring 87.7% on MedQA, within 3 points of DeepSeek R1 at one tenth the inference cost.

Google DeepMind | Models and APIs | Published

SOURCE Google DeepMind, MedGemma: Our most capable open models for health AI development

Sentence Transformers

Sentence Transformers is transitioning from the UKP Lab at TU Darmstadt to Hugging Face, with over 16,000 models on the Hub serving more than a million monthly unique users, and Tom Aarsen continuing as maintainer.

Hugging Face | Models and APIs | In Edition 6,

SOURCE Hugging Face, Sentence Transformers is joining Hugging Face

UK data residency

OpenAI is introducing UK data residency on Friday 24th October for API Platform, ChatGPT Enterprise, and ChatGPT Edu customers, and a new agreement with the UK Ministry of Justice will provide 2,500 employees with ChatGPT Enterprise access.

OpenAI | Models and APIs | In Edition 6,

SOURCE OpenAI, The next chapter for UK sovereign AI

VirusTotal scanning

Every one of the 2.2M+ public model and dataset repositories on the Hugging Face Hub is being continuously scanned with VirusTotal, comparing file hashes against its threat-intelligence database without sharing raw file contents.

Hugging Face | Models and APIs | In Edition 6,

SOURCE Hugging Face, Hugging Face and VirusTotal collaborate to strengthen AI security

AI for Food Allergies

Hugging Face released Awesome Food Allergy Datasets, the first open collection of datasets on food allergies, to accelerate AI research.

Hugging Face | Models and APIs | Published

SOURCE Hugging Face, AI for Food Allergies

Agent Skills

Agent Skills are folders with instructions, scripts, and resources that Claude loads when relevant, available across Claude apps, Claude Code, and API, with a new /v1/skills endpoint for versioning.

Anthropic | Models and APIs | Published

SOURCE Anthropic, Introducing Agent Skills

Show the other 10 rows

Tools and products

Claude for Financial Services

Anthropic expanded Claude for Financial Services with an Excel add-in, new connectors, and 6 Agent Skills, topping the Finance Agent benchmark at 55.3% accuracy.

Anthropic | Tools and products | Published

SOURCE Anthropic, Advancing Claude for Financial Services

Microfrontends preview links

Preview links between microfrontends projects now serve all routes in the group, eliminating 404s and automatically linking deployments built from the same commit or branch.

Vercel | Tools and products | In Edition 6,

SOURCE Vercel, Preview links between microfrontends projects now serve all paths

Sandbox extendTimeout

Vercel Sandbox now has an extendTimeout method to extend a running sandbox beyond its initial timeout, with Pro and Enterprise plans supporting up to 5 hours and Hobby up to 45 minutes.

Vercel | Tools and products | In Edition 6,

SOURCE Vercel, Dynamically extend timeout of an active Sandbox

Server Actions firewall

Vercel Firewall and Observability Plus now support custom rules targeting specific Server Action names, such as rate limiting app/auth/actions.ts#getUser to 100 requests per minute per IP address, starting with Next.js 15.5.

Vercel | Tools and products | In Edition 6,

SOURCE Vercel, Manage Next.js Server Actions in the Vercel Firewall

Turbo build machines

Turbo build machines are available for all paid plans with 30 vCPUs and 60GB of memory, aimed at Turbopack builds and large monorepos that run tasks in parallel.

Vercel | Tools and products | In Edition 6,

SOURCE Vercel, Faster builds with Turbo build machines

Trace Drains on Marketplace

Vercel Drains now send traces and logs from projects to Marketplace observability providers including Braintrust, Dash0, Statsig, and Kubiks, with more providers coming soon.

Vercel | Tools and products | In Edition 5,

SOURCE Vercel, Introducing Trace Drains on the Vercel Marketplace

Claude Haiku 4.5 on AI Gateway

Vercel's AI Gateway now offers Claude Haiku 4.5 with no other provider accounts required, matching Sonnet 4's performance at lower cost and faster speeds.

Vercel | Tools and products | Published

SOURCE Vercel, Claude Haiku 4.5 now available in Vercel AI Gateway

ChatGPT apps on Vercel

Vercel now supports building and deploying ChatGPT apps with Next.js, using MCP natively in the OpenAI sandbox.

Vercel | Tools and products | Published

SOURCE Vercel, ChatGPT apps support on Vercel

uv Python builds

Vercel now uses uv as the default package manager for all Python builds, making builds 30-65% faster and adding support for uv.lock and pyproject.toml dependency files.

Vercel | Tools and products | In Edition 3,

SOURCE Vercel, Python package manager uv is now available for builds

Robotics, hardware and chips

Waymo all-weather Driver

Waymo is advancing its Driver for snowier weather, having amassed tens of thousands of miles in snowy conditions and using a 6th-generation Driver informed by over 100 million autonomous miles.

Waymo | Robotics, hardware and chips | Published

SOURCE Waymo, Creating an all-weather Driver

Waymo in London

Waymo intends to launch its autonomous ride-hailing service in London in 2026, with the Waymo Driver having driven over 100 million miles.

Waymo | Robotics, hardware and chips | Published

SOURCE Waymo, Hello London! Your Waymo ride is arriving

AMD GPU partnership

OpenAI and AMD announced a 6 gigawatt agreement across multiple generations of AMD Instinct GPUs, with the first 1 gigawatt deployment of MI450 GPUs starting in the second half of 2026.

OpenAI | Robotics, hardware and chips | In Edition 4,

SOURCE OpenAI, AMD and OpenAI announce strategic partnership to deploy 6 gigawatts of AMD GPUs

Breakthroughs

Gemini astronomy

Gemini achieved 93% accuracy across three astronomical datasets by learning from just 15 annotated examples per survey, and improved to 96.7% on MeerLICHT through an iterative human-in-the-loop process.

Google Research | Breakthroughs | In Edition 6,

SOURCE Google Research, Teaching Gemini to spot exploding stars with just a few examples

Google Earth AI

Google Earth AI introduces new Remote Sensing Foundations and Population Dynamics Foundations models, and a Gemini-powered Geospatial Reasoning agent that achieved 0.82 accuracy on a Q&A benchmark versus 0.50 for Gemini 2.5 Pro.

Google Research | Breakthroughs | In Edition 6,

SOURCE Google Research, Google Earth AI: Unlocking geospatial insights with foundation models and cross-modal reasoning

Private synthetic albums

Google Research introduced a method for generating differentially private synthetic photo albums using an intermediate text representation and hierarchical generation, tested on the YFCC100M dataset.

Google Research | Breakthroughs | In Edition 6,

SOURCE Google Research, A picture's worth a thousand (private) words

Quantum Echoes

Google Quantum AI demonstrated a verifiable quantum advantage with the Quantum Echoes algorithm on the Willow chip, where experiments took approximately 2 hours versus an estimated 13,000 times longer on a classical supercomputer.

Google Research | Breakthroughs | In Edition 6,

SOURCE Google Research, A verifiable quantum advantage

DeepSomatic

DeepSomatic identifies tumor variants with higher accuracy, scoring 90% F1 on Indels with Illumina data versus 80% for the next-best method.

Google Research | Breakthroughs | Published

SOURCE Google Research, Using AI to identify genetic variants in tumors with DeepSomatic

Funding and acquisitions

OpenAI acquires Sky

OpenAI acquires Software Applications Incorporated, maker of Sky, a natural language interface for the Mac.

OpenAI | Funding and acquisitions | Published

SOURCE OpenAI, OpenAI acquires Software Applications Incorporated, maker of Sky

YC Visiting Partners

YC announced nine new Visiting Partners, including Harshita Arora, the youngest ever, and founders from Swiftype, Dependabot, and Cron.

Y Combinator | Funding and acquisitions | Published

SOURCE Y Combinator, Meet YC's Newest Visiting Partners

One line every time AI actually ships, around the clock, free on Threads and Bluesky. Every one is in The Deploy Log with its source.

The Saturday edition is the week in one, free to read on the site, and by email if you want it: get it by email.