AI × GTM News

Perplexity is integrating OpenAI's GPT-6 Astra to automate software development, internal communications, and production monitoring.

OpenAIPerplexity

LangChain introduced Connections to securely manage user credentials and OAuth for agents, allowing them to act with specific per-caller identities.

LangChain

Cognition is utilizing OpenAI's GPT-6 Astra to enhance Devin's software testing capabilities and reduce manual code review for engineers.

OpenAICognition

Together AI updated its fine-tuning platform with additional open-weight models, live metrics, Expert LoRA, and lower training prices.

Together AI

Vercel has doubled the storage capacity for all Vercel Sandboxes to 64 GB to accommodate larger repositories, dependencies, and storage-intensive AI agent tasks.

Vercel

Pro and Enterprise teams can now restrict the creation and management of Vercel Connect connectors to specific roles or users with extended permissions.

Vercel

Clay secured $115M to expand its AI growth engine platform following 4x revenue growth and adoption by major AI companies including OpenAI and Anthropic.

Clay

Vercel now supports the OpenAI Agents API, providing hosted infrastructure and Sandboxes for code execution to power long-running, tool-using AI agents.

VercelOpenAI

The new Data agent allows ChatGPT Work users to connect company data, generate insights, and build interactive dashboards using natural language.

OpenAI

OpenAI launched a specialized version of ChatGPT for financial services that integrates built-in financial data with GPT-6 Astra for research and modeling.

OpenAI

OpenAI introduced the Agents API, a managed service powered by the Codex harness designed for building long-running AI agents and tool orchestration.

OpenAI

Vercel's AI SDK harness layer now supports GitHub Copilot, allowing developers to switch between different coding agents using a unified interface.

VercelGitHub

Clay released an updated version of its advertising product designed to integrate account-based data with ad platform delivery.

Clay

Sierra released next-generation multimodal agents that integrate voice, text, and visuals within a single customer conversation interface.

Sierra

Cursor Projects introduces a coordinator agent that plans and delegates tasks to subagents to manage large-scale features and migrations over time.

Cursor

OpenAI is partnering with the GSA to provide government agencies with $0 license fees, discounted usage, and enhanced cyber defense support.

OpenAIGSA

Clay released an updated version of its outbound sequencing tool to enhance automated sales outreach workflows.

Clay

FastAPI frontends and static files are now automatically promoted to the Vercel CDN at build time, reducing latency by avoiding function invocation.

Vercel

Clay updated its Claygent AI agent to support new open weight models, providing users with more flexibility in model selection for web research.

Clay

Together AI launched preemptible compute options for GPU Clusters, offering the same capacity at 50% of the on-demand rate with a five-minute drain window.

Together AI

Cloudera and Mistral AI have partnered to integrate Mistral's models into Cloudera's platform for regulated industries.

Mistral AICloudera

Replit has released native Databricks Lakebase support to allow enterprise teams to utilize governed data for production-ready applications.

ReplitDatabricks

Together AI optimized its ThunderKittens framework for NVIDIA's Vera Rubin NVL72 hardware, achieving over 22 PFLOPS performance.

Together AINVIDIA

AiSDR introduced Ami, an AI-powered go-to-market agent designed to automate lead selection, messaging strategy, and campaign optimization.

Writer introduced Enterprise Brain, a shared context layer designed to store company-specific facts and patterns for GTM teams using its AI platform.

Writer

Pinecone has moved its Full-Text Search feature to general availability, integrating BM25 keyword ranking and Lucene syntax with dense vector indexes.

Pinecone

IBM has released a state-of-the-art time series model with a commercial-friendly license, hosted on Hugging Face.

Pinecone introduced a new Documents API that allows users to model text, dense, and sparse fields within a single schema to simplify data management.

Pinecone

Lovable launched a partner program for experts and solution providers to build applications for clients using its platform.

Mistral AI secured €3 billion in Series D funding at a €21 billion valuation to develop its sovereign, open-weight AI technology.

Mistral AI

AlphaGenome Atlas is a predictive map covering 9 billion single-letter DNA variants across the human genome to assist in genomic research.

Sierra released hyper-tau-bench, an open-source framework for evaluating agents that possess the ability to construct other agents over long horizons.

Sierra

LangChain added context modes to its deepagents framework, allowing subagents to either fork supervisor context or operate in isolation for efficiency.

LangChain

Lovable introduced a drafts feature allowing teams to experiment with multiple versions of a project without affecting the live application.

LangChain has added native MCP support via the langchain.mcp library, integrating FastMCP for stateless protocol handling and tool caching.

LangChain

A new efficient multimodal-native and multilingual encoder model has been released on Hugging Face.

AiSDR released a comparative positioning analysis between its new AI agent, Ami, and Artisan’s AI SDR, Ava, focusing on cost-efficiency and time savings.

WeatherNext 3 is a new global weather forecasting AI model designed for improved accuracy in meteorological predictions.

HubSpot introduced HubSpot AEO, a tool for continuous tracking and prioritized recommendations to manage brand visibility across AI assistants and answer engines.

HubSpot

Google DeepMind expands its model lineup with Gemini 3.8 Flash and a specialized version for cybersecurity applications.