Claude Sonnet 5 and Claude Science Put Agents to Work

In partnership with

Cut Lead Review From Hours To Minutes

Sign up for a free trial of Attio, the agentic CRM.

Ask Attio to build a daily workflow that surfaces the deals that need your attention today, like anything with a stage change, a recent reply, or a new signal in the last 24 hours.

Review your pipeline in Claude, synced live from Attio via MCP.

That's it.

Try Attio Now

Claude’s Workhorse Goes Agentic

Anthropic launched Claude Sonnet 5 as its most agentic Sonnet model, emphasizing planning, browser and terminal tool use, coding, and autonomous execution. The positioning is clear: bring Opus-like capability closer to high-volume enterprise workloads without forcing every agent task onto the most expensive model tier.

The model is available across Claude plans, Claude Code, and the Claude Platform, with introductory API pricing of $2 per million input tokens and $10 per million output tokens through August 31 before moving to $3/$15. Anthropic says Sonnet 5 improves on Sonnet 4.6 in agentic search and computer-use evaluations, while adding stronger safeguards for prompt injection, hallucination, and cyber misuse.

Strategically, this pushes the agent market toward cheaper, more capable “daily driver” models that can handle real work across browsers, terminals, and codebases.

Claude Turns Labs Into Agent Workbenches

Anthropic introduced Claude Science, an AI workbench designed to help scientists run literature review, analysis, computation, and research workflows through agentic systems. The platform includes more than 60 curated skills and connectors across areas like genomics, proteomics, structural biology, cheminformatics, and single-cell analysis.

Claude Science can spawn specialist agents, ask for permission before accessing new resources, and use a reviewer agent to check citations, calculations, and reproducibility. Anthropic is also backing projects with up to $30,000 in credits plus $2,000 in Modal compute, with applications open through July 15.

The strategic implication is big: agent platforms are moving from general productivity into domain-specific scientific execution, where auditability and reproducibility matter as much as autonomy.

NVIDIA Gives Science Agents a GPU Toolkit

NVIDIA announced that its BioNeMo Agent Toolkit is integrated into Claude Science, giving research agents access to NVIDIA-accelerated biology and chemistry tools. The toolkit packages capabilities such as Evo 2, Boltz-2, OpenFold3, RAPIDS-singlecell, nvMolKit, and Parabricks as callable skills for scientific agents. NVIDIA says 18 of the top 20 pharma companies already use BioNeMo, giving the integration immediate credibility in enterprise life sciences.

Performance claims include Parabricks cutting workflows from hours to minutes, RAPIDS-singlecell reducing a 1.3-million-cell workflow from 52 minutes to 25 seconds, and nvMolKit delivering up to 3,000x speedups.

Strategically, this shows the agent stack splitting into two layers: generalist reasoning agents on top, domain-specific accelerated toolkits underneath.

TikTok Opens Ads to Agent Workflows

TikTok launched TikTok Agentic Hub, a marketplace that connects AI agents to TikTok Ads workflows for campaign creation, creative generation, catalog management, and performance analysis. The hub is built around TikTok for Business MCP, allowing AI agents to securely interact with TikTok Ads capabilities.

TikTok says marketers can use ready-made skills, custom skills, or connect agents directly without dealing with API credentials, coding, or complex setup. Launch partners include companies such as HubSpot, Wix, Constant Contact, WorkMagic, Innovid, Kochava, Shoplazza, Mobvista, and AI Rudder.

Strategically, TikTok is turning ad operations into an agent-accessible surface, making MCP a practical bridge between marketing platforms and autonomous campaign workflows.

Klaviyo Links Marketing and Service Agents

Klaviyo announced that Composer, its AI marketing agent, is moving into public beta while its Customer Agent is expanding across service workflows. The key architecture is a shared real-time customer profile: service-agent interactions update preferences, interests, and intent, while Composer uses that same data to build campaigns.

Composer can identify revenue opportunities, build email and SMS campaigns, and draw from patterns across nearly 200,000 brands. Customer Agent supports chat, email, SMS, and WhatsApp, with capabilities such as returns, loyalty-point actions, multilingual support, and daily self-checks.

Strategically, Klaviyo is positioning agents not as isolated bots but as a connected CRM loop where service intelligence feeds marketing execution.

The Ultimate Guide for Usage-Based Pricing for SaaS and AI

Implementing usage-based pricing successfully requires more than just a pricing strategy.

Download this guide for practical advice and best practices when considering usage-based pricing. 

👉 Get your guide

🗺️ The Map of AI Agents (Live & Growing)

We’re charting the entire AI agent ecosystem — thousands of options across categories.
Your next agent is already on the map.

👉 Explore the Landscape Map →