Features
Multi-model Beam chats (2-24 models in parallel)
Beam Merge: fuse responses from multiple models
Personas with persistent instructions, memory, custom parameters
Advanced model controls: effort, temperature, verbosity, reasoning budget
AI Inspector: view raw API requests and responses
Automatic retry and provider timeout handling
File attachments: PDF, PPT, Word, Excel, images
Voice input and output (ElevenLabs, OpenAI TTS, Web Speech)
Web browsing and search (Google Search API)
Image generation (DALL·E, GPT image)
Code highlighting and execution
Persona library with hundreds of presets
On-premise or cloud deployment
1M context window for supported models (e.g., Claude Mythos)
Generative image creation via API
Image editing (extend, remove, fill) APIs
Custom model fine-tuning (enterprise)
20+ generative and creative APIs
SDKs for Node.js, Python, Java
Enterprise-grade security compliance (SOC2, ISO27001, FedRAMP)
Usage-based pricing (pay-as-you-go)
Scalable cloud infrastructure (AWS)
API collections for common workflows
Integration with Adobe Experience Cloud (AEM, Workfront)
Commercial safety (licensed training data)
Content credentials (Adobe C2PA integration)
Developer portal with documentation
Rate limiting and access controls
