The wait is over. Since the debut of ChatGPT, no release has been more anticipated.
Introducing GPT-5, now available in Saga.
What makes GPT-5 stand out is not just its leap in intelligence and accuracy, but its unique ability to understand the complexity of your question, and deciding both the level of power required and how long it should think and reason before answering. By combining the strengths of traditional Large Language Models with advanced Reasoning Models, GPT-5 goes beyond being just a model, it’s a complete system that offers a far more intuitive way to work with AI.
Saga will now use GPT-5 as the default model/system, while also offering GPT-5 mini as a more sustainable, smaller model. You no longer need to manually switch to a reasoning model — a process that often overcomplicated simple tasks like summarization — and still get exactly the right level of reasoning for your question.
Although we love training legal professionals in AI, eliminating the need to explain the difference between classical LLMs and reasoning models will make the adoption journey even better.
Redefining Advanced Agentic Workflows
The arrival of GPT-5 marks a new inflection point for the legal industry. Beyond improvements in language understanding, instruction following, and variable reasoning, GPT-5 is amazing at tool calling, therefore making it possible to autonomously plan, execute, and adapt complex sequences of tasks.
By combining these with agentic capabilities powered by GPT-5, Saga evolves into an orchestrated legal intelligence system.
In practice, this means legal teams can:
- Create AI agents that interact seamlessly with legal databases, Projects, DMS systems, and client-specific knowledge bases.
- Delegate repetitive, research-heavy, or procedural tasks to AI agents that follow clearly defined policies and quality benchmarks.
- Automatically surface, cross-verify, and format relevant information for faster decision-making.
Human-Centric AI by Design
A defining principle of Saga’s approach is that AI must serve people. Human-centric AI ensures that lawyers remain in control of processes, with systems built to be transparent, explainable, and aligned with professional ethics.
Within Saga’s agentic workflows, GPT-5 acts as a powerful orchestrator — not a black box. Every action is traceable, outputs can be verified, and legal professionals can intervene, adjust, or redirect the AI at any stage. This safeguards both accuracy and trust, while fostering user confidence in the technology.
One of the other significant improvements are GPT-5’s citation capabilities, reducing the risk of hallucinations and ensuring legal professionals can validate the AI’s work at all times.
Designing for the Future
With GPT-5 as the engine and Saga as the legal-specific infrastructure, advanced agentic architectures become a practical, scalable reality. The future of legal practice will not be defined solely by adopting new technology, but by designing systems that combine human judgment, intelligent AI, and transparent automation. GPT-5, working through Saga’s platform, is a decisive step in that direction.