GenAI PMOpus 4.5
A model used to power v0 Max in the newsletter. For AI PMs, it signals model selection as a product differentiation and cost lever.
Key Highlights
- Opus 4.5 appears across coding, prototyping, and browser-agent products as a high-capability model layer.
- v0 Max used Opus 4.5 as a default-model upgrade while also lowering price, showing model choice as a business lever.
- Teams used Opus 4.5 in practical workflows with Claude Code, Cursor, Figma MCP, and internal prototyping stacks.
- The newsletter examples show Opus 4.5 often works best with process guardrails like plan mode, custom skills, and shared context files.
- Opus 4.5 is a useful case study for AI PMs deciding how to balance capability, autonomy, reliability, and cost.
Opus 4.5
Overview
Opus 4.5 is a frontier model referenced across coding, browser automation, and prototyping workflows in the newsletter ecosystem. It appears as the model powering products like v0 Max, Claude Code-based workflows, and agentic tools such as Comet and Factory’s Droid. In practice, Opus 4.5 shows up less as a standalone destination product and more as the intelligence layer behind higher-level developer and product-building experiences.For AI Product Managers, Opus 4.5 matters because it illustrates a core product strategy pattern: model selection itself can be a differentiator. Across the mentions, teams use Opus 4.5 to improve prototype quality, enable higher-autonomy coding flows, and position premium experiences like “Max” tiers. It also appears directly tied to pricing and packaging decisions, as seen when v0 Max became the default while also getting cheaper. That makes Opus 4.5 relevant not just as a capability choice, but as a lever for UX quality, cost control, and go-to-market positioning.
Key Developments
- 2026-01-17: v0 announced that v0 Max became the new default for all users, powered by Opus 4.5, while also becoming 20% cheaper. This is a clear example of model choice being used as both a product-quality and pricing lever.
- 2026-01-22: Comet adopted Opus 4.5 as the default model for its browser agent for all Max subscribers, showing the model’s role in premium agent experiences.
- 2026-01-24: In a Claude Cowork / Claude Code workflow demo, Boris recommended starting sessions in plan mode, then switching to auto-accept edits with Opus 4.5 for one-shot code generation. He also emphasized maintaining a shared Cloud.md knowledge base in Git for better team guidance.
- 2026-01-31: Dharmesh Shah highlighted a reliability issue with agentic coding assistants and recommended creating custom skills—reusable instructions and context—to reduce mistakes and improve consistency when using advanced models like Opus 4.5.
- 2026-02-16: Peter Yang used Factory’s Droid agent via the Ghosty CLI in high-autonomy spec mode with Opus 4.5 for planning and GPT-5.2 for execution to build and QA a React speed-reading app. This showed a multi-model workflow where Opus 4.5 was chosen for planning rather than all tasks.
- 2026-02-24: Notion AI’s design team was described as using a Next.js monorepo called Prototype Playground, powered by Claude Code (Opus-4.5) and Cursor, to rapidly create production-ready prototypes with AI-assisted tooling.
- 2026-03-23: Felix Lee demonstrated using Claude Code (Opus 4.5) inside Cursor with Figma MCP to turn a Figma landing page into production-ready code in under 15 minutes, including refinements like form states, a fixed logo, and a sticky PDF component.
Relevance to AI PMs
1. Use model selection as a product and pricing lever. The v0 Max example shows that choosing a stronger model can support both better default UX and packaging changes such as premium tiers or cost reductions. PMs should treat model choice as part of the product surface, not just an infrastructure detail.2. Design workflows around task-model fit. The Factory example shows Opus 4.5 being used specifically for planning while another model handled execution. PMs can improve reliability and cost efficiency by assigning different models to planning, generation, QA, and tool use instead of forcing a single-model architecture.
3. Pair strong models with process guardrails. The Claude Code and Dharmesh Shah examples suggest that advanced models still need scaffolding: plan mode, reusable skills, team knowledge files, and explicit operating instructions. PMs shipping agentic experiences should define these guardrails early to reduce hallucinations, inconsistency, and rework.
Related
- claude-code: One of the main contexts in which Opus 4.5 appears; used for coding and agentic software tasks.
- cursor: Frequently paired with Claude Code (Opus 4.5) for IDE-based code generation and editing workflows.
- figma-mcp: Connected through design-to-code workflows, where Opus 4.5 helps convert Figma designs into implementation.
- notion / prototype-playground: Examples of internal product and design prototyping environments powered by Claude Code with Opus 4.5.
- factory / droid / ghosty-cli: Illustrate higher-autonomy agent workflows where Opus 4.5 is used for planning in a multi-model stack.
- gpt-52: Mentioned alongside Opus 4.5 in a split workflow, reinforcing the idea of using different models for different stages.
- cloudmd / claude-cowork: Related to team guidance and operating patterns for getting more reliable results from Opus 4.5-powered agent sessions.
- comet / v0: End-user products where Opus 4.5 is surfaced as the default model behind premium or upgraded experiences.
- dharmesh-shah / aravind-srinivas: People associated with practical usage patterns and product decisions involving Opus 4.5.
Newsletter Mentions (7)
“Felix Lee demonstrates using Claude Code (Opus 4.5) in the Cursor IDE with Figma MCP to convert a Figma landing page design into production-ready code in under 15 minutes.”
#1 ▶️ How to Design and Code with Claude Code and Figma MCP in 50 Min | Felix Lee Peter Yang Felix Lee demonstrates using Claude Code (Opus 4.5) in the Cursor IDE with Figma MCP to convert a Figma landing page design into production-ready code in under 15 minutes. Implemented a Figma landing page via Claude Code within Cursor and Figma MCP, adding form submit states, fixing a missing logo, and making a PDF “book” component sticky—all in just 15 minutes.
“Notion AI’s design team uses a Next.js monorepo called Prototype Playground, powered by Claude Code (Opus-4.5) and Cursor, to rapidly build production-ready prototypes with AI-assisted tooling.”
GenAI PM Daily February 24, 2026 GenAI PM Daily 🎧 Listen to this brief 3 min listen Today's top 23 insights for PM Builders, ranked by relevance from Blogs, YouTube, X, and LinkedIn. OpenAI Updates SWE-bench Verified Metrics #1 📝 OpenAI News Why SWE-bench Verified no longer measures frontier coding capabilities - OpenAI explains why the SWE-bench Verified benchmark is no longer used to measure frontier coding capabilities, outlining limitations of the metric and reasons it can misrepresent real-world model performance. The piece describes the rationale for retiring or deprioritizing the benchmark and points toward alternative evaluation approaches for assessing coding ability. Also covered by: @Sebastian Raschka #2 📝 Simon Willison Ladybird adopts Rust, with help by AI - Andreas Kling describes using coding agents (Claude Code and Codex) to port Ladybird's LibJS JavaScript engine from C++ to Rust, producing byte-for-byte identical output and completing ~25,000 lines of Rust in about two weeks.
“Peter Yang Uses Factory’s Droid agent via the Ghosty CLI in high-autonomy spec mode with Opus 4.5 for planning and GPT-5.2 for execution to build and QA a React-based speed-reading web app using Chrome DevTools for automated screenshots, linting and type-checking.”
#3 ▶️ Full Tutorial: The Most Underrated AI Agent for Coding and Product Work | Eno Reyes (Factory) Peter Yang Uses Factory’s Droid agent via the Ghosty CLI in high-autonomy spec mode with Opus 4.5 for planning and GPT-5.2 for execution to build and QA a React-based speed-reading web app using Chrome DevTools for automated screenshots, linting and type-checking.
“To mitigate this, he recommends defining custom “skills”—preconfigured instructions and context—for frequent tasks.”
From LinkedIn • Deeper Insights AI Tools & Applications Dharmesh Shah’s post highlights a common reliability pitfall with agentic coding assistants like Claude Code: when they reply “you’re absolutely right,” they may overlook errors or violate principles such as DRY (Don’t Repeat Yourself). To mitigate this, he recommends defining custom “skills”—preconfigured instructions and context—for frequent tasks. This approach helps maintain consistency and reduce surprising mistakes, even when leveraging advanced models like Opus 4.5.
“Boris recommends starting Claude Code sessions in plan mode to refine a task, then switching to auto-accept edits with Opus 4.5 for one-shot code generation, and maintaining a shared Cloud.md (checked into Git) with a GitHub action that lets you tag the Claude app on PRs to automatically update team guidance.”
I got a private lesson on Claude Cowork & Claude Code Greg Isenberg • January 23, 2026 In this episode, Boris demonstrates how Claude Cowork’s simple Mac app—powered by the same Claude Code “quad agent”—lets users mount local folders to rename receipt files, generate and format spreadsheets in Chrome, and send emails via Gmail, and shares best practices for Claude Code sessions using plan mode, Opus 4.5, and a shared Cloud.md knowledge base. Key Takeaways: Claude Cowork is a Mac-only desktop app (Windows coming soon) that uses the Claude Code “quad agent” SDK to access user-mounted folders, rename files by date with reverse elicitation, and output new files like spreadsheets. With the Cowork Chrome extension, the agent automates browser tasks—creating and formatting a Google Sheet from receipt data and composing an email in Gmail—while enforcing safety via a sandbox VM and deletion protection.
“Comet Defaults to Opus 4.5 : Aravind Srinivas @AravSrinivas announced the adoption of Opus 4.5 as the default model for the browser agent on Comet for all Max subscribers .”
Google Updates GeminiApp UX with Instant Answers From X AI Product Launches & Updates Comet Defaults to Opus 4.5 : Aravind Srinivas @AravSrinivas announced the adoption of Opus 4.5 as the default model for the browser agent on Comet for all Max subscribers . GeminiApp UX Improvement : Josh Woodward @joshwoodward fixed a papercut by adding an "Answer now" button to stop GeminiApp from thinking and deliver instant responses using the 3 Flash model . World API Launch : Fei-Fei Li @drfeifei shared that the World API is live, enabling Marble to empower user creations, products, and workflows.
“v0 Max default powered by Opus 4.5 : V0 @v0 announced v0 Max as the new default for all users, powered by Opus 4.5 and now 20% cheaper .”
AI Tools & Applications v0 Max default powered by Opus 4.5 : V0 @v0 announced v0 Max as the new default for all users, powered by Opus 4.5 and now 20% cheaper . Session renaming & connectors in research preview : Claude @claudeai shipped session renaming , connector improvements , and early feedback fixes in its macOS app research preview . AWS database integration : V0 @v0 introduced AWS databases support, enabling a prompt-to-app workflow with a production database in minutes and no manual setup .
Related
Anthropic's coding assistant used for programming and automation tasks. The newsletter references it for building a custom approval device and for writing and research workflows inside AI agents.
An AI coding editor and automation platform. The newsletter highlights multi-repository support for automations across codebases.
Co-founder and CTO of HubSpot. He is associated here with launching HubSpot's Agent CLI and advocating human-agent collaboration.
The Perplexity founder/CEO, mentioned discussing enterprise adoption, security engineering, and Perplexity Computer. He appears here as a voice on agentic security workflows and search infrastructure.
A UI/product-building tool that now includes an automatic fix for pull request conflicts. The feature uses an AI agent to merge and resolve base-branch conflicts.
Anthropic's collaborative AI tool used for multimodal workflows, code execution, and connector-based access to external data sources. It appears in the newsletter as a practical example of an AI assistant handling planning, analysis, and automation tasks.
A GPT model release referenced as an impressive model by Kevin Weil. For AI PMs, it represents continued frontier-model iteration and user expectation growth.
A productivity company referenced through the Notion AI agent Hot Potato. It appears here as the host context for an internal standup-prep automation.
A plugin that enables code-to-design roundtrips in Figma. It is relevant as an interoperability layer between AI-generated code and design tooling.
A standalone browser from Perplexity designed to let a personal-computer AI execute web tasks reliably.
An AI-native startup mentioned as delegating tasks to AI agents across multiple functions. Relevant to PMs as an example of an AI-first operating model.
Stay updated on Opus 4.5
Get curated AI PM insights delivered daily — covering this and 1,000+ other sources.
Subscribe Free