Foundry Buys Griptape to Wire AI Into VFX Pipelines

The VFX software giant acquired the enterprise orchestration platform to give studios control over multiple AI models within existing production workflows.

Foundry completed its acquisition of Griptape on February 18, positioning the VFX software company to build secure AI orchestration directly into tools like Nuke and Katana. The deal brings Griptape's Python-based framework, designed by AWS veterans to manage multiple AI models with enterprise-grade security, into Foundry's ecosystem at a moment when studios are moving from AI experiments to daily production use.

The acquisition addresses a specific pain point: studios need to orchestrate video diffusion models, LLMs, and other AI tools while maintaining the strict provenance and security requirements of production pipelines. According to Digital Production, Griptape offers a dual-layer architecture combining a Python framework with a node-based interface, allowing technical directors to build custom AI agents that can be deployed safely across teams.

"We're empowering storytellers with efficient, controlled AI workflows," Foundry CEO Jody Madden told Broadcast, emphasizing that artists would retain control while leveraging automation. The company declined to disclose financial terms.

The timing follows Foundry's 2021 release of CopyCat, its machine learning-powered rotoscoping tool, suggesting a gradual build toward what the company now calls an AI-first pipeline strategy. Where CopyCat focused on a single task, Griptape enables orchestration across multiple models. This matters as studios increasingly combine video generation, texture synthesis, and other AI capabilities within single shots.

Griptape's architecture appears designed for the specific constraints of VFX production. The platform provides what Foundry describes as traceability, the ability to track which models generated which assets, when, and with what parameters. This addresses legal and artistic concerns about AI-generated content mixing unmarked into final deliverables.

The framework also handles model switching, allowing facilities to use open-source models for concept work while routing final shots through commercial APIs with clearer licensing. According to Animation World Network, this flexibility lets studios manage costs while maintaining quality standards across different production phases.

For smaller studios and independent creators, the acquisition could democratize access to enterprise-grade AI orchestration. Griptape's Python foundation means technical artists familiar with pipeline scripting can build custom workflows without learning proprietary languages.

Studios gain traceable AI orchestration within existing Foundry tools rather than separate platforms. The Python-based framework allows custom agent creation without vendor lock-in. The dual architecture supports both node-based artist workflows and code-based TD customization. Model-agnostic design enables mixing open-source and commercial AI services. Security features address studio concerns about data leakage and asset provenance.

The integration timeline remains unspecified, though Foundry indicated Griptape's tools would complement rather than replace existing features. The company plans to maintain Griptape's model-agnostic approach, suggesting studios won't be forced into specific AI vendors. This matters as the video generation field rapidly evolves with new models appearing monthly.