GPT-5.1 Launched | Baidu Releases ERNIE-4.5-VL-28B-A3B-Thinking

Time is limited, so we will be direct. Here is your newsletter AI Dev Brief from Marktechpost, covering key research, models, infra tools, and practical updates for AI developers and researchers

GPT-5.1 is an in generation upgrade to OpenAI’s GPT-5 stack that introduces GPT-5.1 Instant and GPT-5.1 Thinking, both with adaptive reasoning so the models spend more compute on hard prompts and less on simple ones, improving math and coding benchmarks while keeping latency low. GPT-5.1 adds explicit personalization with preset styles and tone sliders, and reuses the GPT-5 safety framework with updated system card metrics that raise safety scores for gpt-5.1-instant and strengthen jailbreak robustness, making ChatGPT more controllable for production engineering workflows. Read the full launch insights/article here.

Maya1 is a 3B parameter, decoder only, Llama style text to speech model that predicts SNAC neural codec tokens to generate 24 kHz mono audio with streaming support. It accepts a natural language voice description plus text, and supports more than 20 inline emotion tags for fine grained control. Running on a single 16 GB GPU with vLLM streaming and Apache 2.0 licensing, it enables practical, expressive and fully local TTS deployment. Read the full launch insights/article here.

Germany based open-source remote access company - NetBird just built an "AI Mega Mesh". A project that started out to prove that multi-cloud networking doesn’t have to be complicated, resulted in creating a secure AI inference infrastructure that connects GPU resources across multiple cloud providers using Microk8s, vLLM, and NetBird. Read the full launch insights/article here.

ERNIE-4.5-VL-28B-A3B-Thinking is Baidu’s new lightweight multimodal reasoning model that activates only 3B parameters while building on a larger ERNIE-4.5-VL architecture, and it targets high accuracy on document, chart and video understanding with features such as “Thinking with Images” for zoom based inspection and tool utilization for image search, all released under Apache License 2.0 for commercial deployment through standard stacks like transformers, vLLM and FastDeploy. Read the full launch insights/article here.

▶ [Open Source] Memori: An Open-Source Memory Engine for LLMs, AI Agents & Multi-Agent Systems Codes & Examples

▶ How to Reduce Cost and Latency of Your RAG Application Using Semantic LLM Caching Codes Tutorial

▶ A Coding Implementation to Build and Train Advanced Architectures with Residual Connections, Self-Attention, and Adaptive Optimization Using JAX, Flax, and Optax Codes Tutorial

▶ How to Build an End-to-End Interactive Analytics Dashboard Using PyGWalker Features for Insightful Data Exploration Codes Tutorial

▶ How to Build a Fully Functional Custom GPT-style Conversational AI Locally Using Hugging Face Transformers Codes Tutorial

100+ more open codes/notebooks here ➡️