Blender pushes back on Anthropic

In partnership with

Welcome to VP Land! Fresh-ish off the NAB show floor, where we hunted down every AI tool worth talking about, we’re back with the newsletter.

In today’s edition:

Anthropic’s creative push hits Blender pushback
The AI standouts we found at NAB 2026
GPT Image 2.0 plans before it draws
Oscars draw a hard line on AI

Anthropic’s Creative Push Hits Blender Wall

Anthropic launched MCP connectors for nine creative applications, opening Claude up to drive Adobe Creative Cloud, Blender, Ableton, Affinity by Canva, Autodesk Fusion, Resolume Arena/Wire, SketchUp, and Splice. The Blender integration became the flashpoint.

— # (#)

To fund the official Blender connector, Anthropic joined as a Patron sponsor at €240,000 per year, roughly four full-time developers. Blender’s existing user base and contributing developers pushed back with donation boycott threats, and Blender CEO Francesco Siddi acknowledged the misstep: “We should have opened up more conversation and perspectives from contributors before making this decision, and for that we are sorry.”

The donation is now one-off. Anthropic’s contribution stands as a single payment rather than the recurring annual commitment. The Blender Foundation says it will “strengthen the processes when it comes to accepting donations.”
No AI features planned for Blender. “No generative AI functionality is currently available or planned to be integrated in Blender,” the Foundation stated, distancing the connector from any in-app AI integration.
What the connector does. The MCP integration lets Claude or any LLM analyze scenes, debug rigs, and batch-apply changes across objects through Blender’s Python API. It works with any LLM, not just Claude.
Adobe’s connector covers 50+ tools. Adobe’s integration gives Claude users natural-language access to 50+ Creative Cloud tools, with the assistant orchestrating workflows behind the scenes.
The dissent came from inside the house. Blender Guru Andrew Price calls the Foundation’s decision a strategic mistake, arguing existing contributors and the wider community should have accepted the recurring funding rather than reject it on principle.

Worth noting: the pushback came from Blender’s own user base and contributors, the people who use the software daily. The Foundation traded ongoing developer funding for community trust.

SPONSOR MESSAGE

Attio is the AI CRM for high-growth teams.

Connect your email, calls, product data and more, and Attio instantly builds your CRM with enriched data and complete context. Whether you’re running product-led growth or enterprise sales, Attio adapts to your unique GTM motion.

Then Ask Attio to plan your next move.

Run deep web research on prospects. Update your pipeline as you work. Find customers and draft outreach emails. Powered by Universal Context, Attio's intelligence layer, Attio searches, updates, and creates across your data to accelerate your workflow.

Ask more from your CRM.

Ask Attio

The AI Standouts From NAB 2026

VP Land spent the week on the NAB 2026 show floor running booth interviews. The AI thread ran through nearly every aisle — these four stood out.

Lightcraft Spark. Browser-based previs where filmmakers plan shots through a unified database synced directly to production.
Cutback Selects. AI assistant editor that takes raw footage to rough cut in minutes instead of a week. We spoke with Cutback CEO Tom Kim at the booth.
Shade. AI auto-tagging for entire video libraries with cloud-based editing and a fully mountable drive that works natively in Premiere, Resolve, and Avid.
Martini. Generative video platform where teams generate images and video, build timelines, step into 3D scenes from a single image, and connect Claude AI agents through MCP.

The takeaway: previs, editing, asset management, and generation all shipped browser-first AI tools at this show. Full interviews are rolling out on the VP Land YouTube channel.

GPT Image 2.0 Adds Thinking Mode

OpenAI explained the architecture behind GPT Image 2.0, framing it as a step beyond single-shot image generation.

— # (#)

Thinking Mode. The model researches the web, plans layouts, and self-checks outputs before generation, rather than producing a one-pass result.
Text accuracy. OpenAI claims roughly 99% text accuracy across Latin, CJK, Hindi, and Bengali scripts. That’s historically the weakest area in image generation.
Resolution and speed. Generations go up to 4K (4096x4096) and run twice as fast as gpt-image–1.5.
Architecture changes. Built on a GPT–5.4 backbone with what OpenAI calls Cognitive Vision Transformer (CVT) architecture and a Physics Inference Layer for spatial reasoning.
Workflow features. Multi-turn editing and Universal Blueprint Mode for reusing layout structures across generations.

What matters: the reasoning pattern that powered GPT–5.4 has now reached image generation, putting layout planning and self-checking inside the model rather than a wrapper around it.

LTX Goes HDR, Kling Hits 4K

Two AI video tools shipped upgrades that move generative video closer to broadcast and theatrical specs.

LTX released an HDR beta, the first AI video model to output HDR instead of 8-bit SDR. The grading implications are significant: highlights don’t clip, shadows don’t crush, and AI footage composites cleanly against CGI plates that already live in HDR. For colorists working with mixed sources, that closes the gap between AI-generated b-roll and primary photography.

Kling AI added one-click 4K generation to Video 3.0, putting native cinematic 4K output inside the model rather than relying on an upscaler. Subscribers get 20% off until May 7.

The takeaway: post-grading and downstream conform have been the practical bottlenecks for AI footage in real productions. HDR support and native 4K close two of the biggest spec gaps holding generative video out of broadcast pipelines.

A mission-driven look at how Project ETERNAL transforms drone-captured 360 footage into immersive, long-lasting 3D archives using Gaussian Splatting to preserve meaningful places.