Fundamentals
Languages
Neutron AI is designed for products that run across many stacks. Use the REST API directly or choose an SDK that fits your backend and agent runtime.
Supported Surfaces
| Environment | Use |
|---|---|
| TypeScript / JavaScript | Web apps, server apps, agent runtimes, provider request builders, and MCP helpers. |
| Python | Automation, notebooks, evaluation workflows, operations tools, and AI applications. |
| PHP | Laravel, Symfony, WordPress, and custom backend integrations. |
| Go | Services, command runners, platform backends, and high-throughput integrations. |
| Rust | High-control services and agent runtimes that prefer strict types and explicit errors. |
| CLI | Developer workflows, CI checks, local testing, support workflows, and scripted operations. |
| REST API | Any environment that can send HTTPS requests. |
| MCP | Model and agent hosts that support remote tool servers. |
Naming Conventions
Public API fields use JSON names such as tenantId, scopeId, and agentId. SDKs may expose idiomatic names for each language while preserving the same request and response contracts.
Provider Independence
Neutron AI does not require one specific model provider. Your app can use the returned context pack with OpenAI, Anthropic, Gemini, Mistral-compatible hosts, private models, or your own routing layer.