Studio
Private AI Workbench for Mac
Free forever with default model — no credit card required.
Product : avgidea.io/en/greenative
Read docs : gn-docs.avgidea.io
Securing Local Data and Code
Greenative Studio allows you to run LLMs in a local environment by integrating with files, tools, and coding agents. By default, it does not send sensitive data to cloud AI services.
No Python, Homebrew, or command-line model configuration is required.
Start local. Keep costs predictable. Connect business data when your team is ready.
Free Plan : Run Phi-4-mini locally — free forever
Pro plan : US $18/month
Apple Silicon Mac (M1 or later) as well as on cloud-based Macs such as Amazon EC2 Macs
Fixed monthly or yearly pricing. No token-based consumption for local processing.
Some tasks are better suited for AI processing in a local environment
While cloud-based AI is powerful, it isn’t necessarily suitable for every task.
Workflows involving business documents, customer data, internal files, and databases require more granular control over where data is processed and stored.
Greenative Studio provides a “local-first” environment for running AI workflows on a Mac.
What you can do
-
Run chats using commercially available local LLMs (Gemma, Phi, Mistral, Nemotron, Qwen) on a Mac with Apple Silicon
-
Work with private files
-
Convert PDFs to PPTX, DOCX, and XLSX
-
Build and test tool-calling workflows
-
Run coding agents on local models
-
Use the bundled MCP server
-
Fine-tune supported models locally using your own data
-
If you need access to business data, connect to the Greenative Platform
Code with Local LLMs
Greenative Studio now includes a local API gateway for agentic coding.
Point your coding agent at your Mac instead of the cloud. Your source code, your prompts, and the model itself stay on your machine.
-
Works with Claude Code and other compatible coding agents
-
One-click start — set host and port, pick a model, done
-
Run coding-focused models like NVIDIA's Nemotron 3.5 Lightning (32 GB+ Apple Silicon)
-
Live token metrics for every session
-
No per-token charges — your agent can code all day
Free plan available — no credit card required.
Coding agent connected to Local LLM through Greenative Studio
Local first by default
Greenative Studio is designed for sensitive work.
Your prompts, files, local workflows, and fine-tuning data do not need to be sent to cloud AI services. External APIs, tools, or MCP servers are only used when explicitly configured.
Fine-tuned models can also stay on your Mac. This helps reduce the risk of exposing not only your source data, but also the model artifacts created from that data.
Greenative Studio supports only commercially usable open-weight models with permissive licenses, such as Apache 2.0, MIT, and OpenMDW, so users can run AI workflows in their local environments with greater clarity.
Running tools locally with Greenative Studio
Manage Gemma, Phi, and Mistral models on the console.
Extend to business workflows
Greenative Studio can be used on its own as a local AI workbench.
When your workflows need access to shared business data, Greenative Studio can connect to Greenative Platform through greenative-data.
Greenative Platform extends Greenative Studio with access to databases, cloud storage, and internal systems.
Available options:
-
SaaS — managed, browser-based access for teams that can securely connect selected data sources
-
Dedicated instances — customer-specific deployments for selected organizations with stricter isolation or governance requirements
FAQ
Can I use Claude Code with Greenative Studio?
Yes, you can.
Launch the Coding Agent gateway in Greenative Studio and set `ANTHROPIC_BASE_URL` to your local endpoint. Requests will always be processed by the model running in Greenative Studio, regardless of the model name specified by the client.
For further details, please refer to the "code" section of the online document.
Can I run Greenative Studio on a cloud-hosted Mac (e.g. Amazon EC2 Mac instances)?
Yes. Your license permits running on Apple Silicon Macs you control, including cloud-hosted instances. All inference happens on that machine — nothing is sent to third-party AI services.
One note for EC2 Mac: stopping and starting an instance may place it on a different physical host, which changes the hardware ID your license is tied to (rebooting does not). If your license stops validating after a stop/start, contact us at support@avgidea.io and we'll transfer it — takes a day or less. A self-service reactivation flow is on the way.
Do I need Greenative Platform to use Greenative Studio?
No. Greenative Studio can be used on its own as a local AI workbench.
Greenative Platform is optional and adds access to shared business data sources such as databases, cloud storage, and internal systems.
Does Greenative Studio send my data to cloud AI services?
Not by default. Greenative Studio is local-first.
Data may be sent outside your Mac only when you configure external APIs, tools, MCP servers, or cloud services.
What is greenative-data?
greenative-data (formerly greenative.mcp) is an MCP server that connects Greenative Studio to Greenative Platform for business data workflows.
Is Greenative Studio a replacement for ChatGPT or Claude?
No. Greenative Studio is not another general-purpose cloud AI chat app.
It is a local-first AI workbench for Mac, focused on private workflows, local models, tool calling, and MCP-based integrations.
What Mac do I need?
Greenative Studio is designed for Apple Silicon Macs (M1 ~ M5). Performance depends on your Mac model, memory size, selected model, and workflow complexity.
Is there a free trial?
Yes. You can try Greenative Studio on your Mac before subscribing.
The free trial is designed to help you check local model performance, supported workflows, and compatibility with your Mac.
