Tag
Together AI 推出 Together Post-Training 服务，让开发者在其平台上通过 RL 与 SFT 定制化训练专属模型，同时由平台负责底层系统与基础设施的协同优化。该服务提供灵活的训练循环定义、独占 GPU 资源（Model Resources）以及 LoRA 与全参数训练支持。
Igor Babuschkin (former DeepMind/OpenAI/xAI) announces River AI, a startup claiming the best system for building internal custom models, backed by a decade of top-tier model training experience and enthusiasm from Anjney Midha.
The tweet emphasizes the need for companies to own their AI intelligence stack, including custom models, harnesses, and evaluations, rather than renting, with a quote from Gabe Pereyra discussing challenges at Harvey.
Fireworks AI's inclusion in trending and fastest-growing categories highlights the growing importance of custom models and owning the AI intelligence stack, with broader trends in AI SaaS vendor growth.
Pallet launches Custom Models for supply chain teams, enabling enterprise AI sovereignty by training dedicated models on proprietary operational data to improve accuracy, reduce costs, and maintain control.
Mira Murati announces Thinking Machines Lab, a new initiative focused on multimodal AI, custom models, and open science, along with a tool called Tinker for training open-weight models.
The article argues that enterprises should post-train their own custom AI models for mission-critical, high-volume use cases to achieve differentiation, cost savings, and control over tradeoffs, rather than relying solely on general frontier models.
AI application layer companies such as Cursor, Decagon, Harvey, and Notion are shifting from using large model APIs to self-trained models. This trend aims to regain control of the data flywheel rather than merely saving costs.
OpenAI introduces improvements to its fine-tuning API with new features including epoch-based checkpoints, comparative playground for model evaluation, third-party integrations, and enhanced dashboard capabilities. The company also expands its custom models program to give developers more control and flexibility in building domain-specific AI solutions.
Harvey, a generative AI platform for legal professionals, partnered with OpenAI to create a custom-trained case law model that reduces hallucinations and improves reasoning for complex legal tasks like document drafting and contract analysis. The custom model, trained on 10 billion tokens of U.S. case law, achieved 97% lawyer preference over standard foundation models.
OpenAI announced the Assistants API, enabling developers to build agent-like experiences with persistent threads, Code Interpreter, Retrieval, and function calling capabilities. The company also launched a Custom Models program for organizations to train domain-specific GPT-4 models with exclusive access.