Kyutai Releases Hibiki-Zero | Exa AI Introduces Exa Instant

Here is your today’s AI Dev Brief from Marktechpost, covering core research, models, infrastructure tools, and applied updates for AI developers and researchers. Also, don’t forget to register for NVIDIA GTC 2026 event (In person/Virtual). NVIDIA has been supporting us to bring free and unlocked AI research and dev news content to you.

Hibiki-Zero is a 3B parameter, decoder-only model designed for simultaneous speech-to-speech (S2ST) and speech-to-text (S2TT) translation that eliminates the need for complex word-level aligned training data. By leveraging a multistream RQ-Transformer architecture and the streaming Mimi audio codec, the system jointly models source audio, target audio, and an "inner monologue" text stream at a 12.5 Hz framerate. The training pipeline first utilizes coarse sentence-level alignments followed by a novel reinforcement learning strategy using Group Relative Policy Optimization (GRPO) and BLEU-based process rewards to optimize the trade-off between translation quality and latency. This approach achieves state-of-the-art results in accuracy, naturalness, and cross-lingual speaker similarity across five language tasks, while demonstrating the ability to adapt to new languages, such as Italian, with less than 1,000 hours of data.....… Read the full analysis/article here.

Exa has launched Exa Instant, a proprietary neural search engine designed to solve the latency bottleneck in AI agent workflows. By bypassing traditional search engine wrappers and using a custom transformer-based stack, Exa Instant delivers web results in under 200ms with network speeds as low as 50ms. This 15x speed improvement allows engineers to treat search as a real-time primitive in RAG pipelines rather than a slow, external dependency. Priced at $5 per 1,000 requests, the model prioritizes semantic intent over keywords, effectively turning the live web into a high-speed context extension for LLMs......… Read the full analysis/article here.

▶ How to Align Large Language Models with Human Preferences Using Direct Preference Optimization, QLoRA, and Ultra-Feedback Codes Tutorial

▶ Meet CopilotKit: Framework for building agent-native applications with Generative UI, shared state, and human-in-the-loop workflows Codes

▶ [In-Depth Guide] The Complete CTGAN + SDV Pipeline for High-Fidelity Synthetic Data Codes Tutorial

▶ A Coding Guide to Design an Agentic AI System Using a Control-Plane Architecture for Safe, Modular, and Scalable Tool-Driven Reasoning Workflows Codes Tutorial

▶ A Coding Implementation for an Agentic AI Framework that Performs Literature Analysis, Hypothesis Generation, Experimental Planning, Simulation, and Scientific Reporting Codes Tutorial

▶ How to Build a Neuro-Symbolic Hybrid Agent that Combines Logical Planning with Neural Perception for Robust Autonomous Decision-Making Codes Tutorial

150+ more open codes/notebooks here ➡️