Tinker Description
Tinker is an innovative training API tailored for researchers and developers, providing comprehensive control over model fine-tuning while simplifying the complexities of infrastructure management. It offers essential primitives that empower users to create bespoke training loops, supervision techniques, and reinforcement learning workflows. Currently, it facilitates LoRA fine-tuning on open-weight models from both the LLama and Qwen families, accommodating a range of model sizes from smaller variants to extensive mixture-of-experts configurations. Users can write Python scripts to manage data, loss functions, and algorithmic processes, while Tinker autonomously takes care of scheduling, resource distribution, distributed training, and recovery from failures. The platform allows users to download model weights at various checkpoints without the burden of managing the computational environment. Delivered as a managed service, Tinker executes training jobs on Thinking Machines’ proprietary GPU infrastructure, alleviating users from the challenges of cluster orchestration and enabling them to focus on building and optimizing their models. This seamless integration of capabilities makes Tinker a vital tool for advancing machine learning research and development.
Tinker Alternatives
Gemini platform is a comprehensive AI development, machine learning, and agent orchestration platform designed to help organizations build and operate intelligent applications at enterprise scale. Developed by Google Cloud as the evolution of Vertex AI, the platform combines foundation model access, autonomous agent development, custom model training, and production deployment infrastructure. Model Garden provides access to more than 200 AI models, including Google's Gemini family, third-party models, and open-source alternatives that developers can evaluate and customize. Agent Studio offers a low-code development environment for designing AI agents, experimenting with prompts, and testing model behavior through a visual interface. The Agent Development Kit enables developers to build sophisticated agents with tool calling, complex reasoning, and multi-agent orchestration capabilities. Google Antigravity provides an agent-powered development environment for coordinating software engineering tasks and automating application development workflows. Machine learning capabilities include custom training, hyperparameter tuning, automated pipelines, model registries, feature management, and monitoring for model performance and drift. Enterprise governance tools provide agent identity management, centralized registration, access controls, security policies, and observability across deployed AI systems. Gemini platform integrates with BigQuery, Google Cloud services, development frameworks, and enterprise applications to support predictive analytics, generative AI, software development, and business process automation.
Learn more about Gemini platform
Hivenet Compute gives you on-demand RTX 5090 and RTX Pro 6000 GPUs, plus vCPU instances, billed per second while they run. Hivenet owns and operates the infrastructure, with data centres in France, the UAE and the US, so EU teams can keep workloads and data in Europe.
Use it for training and fine-tuning, inference, notebooks, rendering and batch jobs. Launch from ready-made templates in a few minutes, or automate everything through the public Compute API. If you'd rather not run the serving layer yourself, the Inference API gives you an OpenAI-compatible endpoint.
Teams get organisations with role-based access, prepaid credits, and S3-compatible object storage alongside compute.
Learn more about Compute with Hivenet
Inkling
Inkling is Thinking Machines’ open-weights foundation model built for customization, multimodal reasoning, and agentic AI workflows. The model uses a Mixture-of-Experts architecture with 975 billion total parameters and 41 billion active parameters, making it large in capacity while activating only a subset of experts per token. Inkling supports up to a 1 million token context window and was pretrained on 45 trillion tokens spanning text, images, audio, and video. It is designed as a broad generalist model with strengths across coding, reasoning, instruction following, factuality, tool use, vision, audio understanding, forecasting, and safety. Developers can tune its thinking effort to trade off latency, cost, and performance, which is useful for production systems that need efficient reasoning at scale. Inkling can be fine-tuned on Tinker, tested in the Inkling Playground, and deployed through partners such as TogetherAI, Fireworks, Modal, Databricks, Baseten, vLLM, SGLang, llama.cpp, and Hugging Face transformers. The model can generate applications, operate tools, create styled artifacts, reason over visual and audio inputs, and support long refinement loops for collaborative work. Thinking Machines also previewed Inkling-Small, a lighter Mixture-of-Experts model with 276 billion total parameters and 12 billion active parameters for lower-cost and lower-latency workloads. By combining open weights, multimodal training, agentic capabilities, efficient reasoning, and fine-tuning support, Inkling gives builders a flexible AI foundation for specialized products and workflows.
Learn more about Inkling
Amazon SageMaker HyperPod
Amazon SageMaker HyperPod is a specialized and robust computing infrastructure designed to streamline and speed up the creation of extensive AI and machine learning models by managing distributed training, fine-tuning, and inference across numerous clusters equipped with hundreds or thousands of accelerators, such as GPUs and AWS Trainium chips. By alleviating the burdens associated with developing and overseeing machine learning infrastructure, it provides persistent clusters capable of automatically identifying and rectifying hardware malfunctions, resuming workloads seamlessly, and optimizing checkpointing to minimize the risk of interruptions — thus facilitating uninterrupted training sessions that can last for months. Furthermore, HyperPod features centralized resource governance, allowing administrators to establish priorities, quotas, and task-preemption rules to ensure that computing resources are allocated effectively among various tasks and teams, which maximizes utilization and decreases idle time. It also includes support for “recipes” and pre-configured settings, enabling rapid fine-tuning or customization of foundational models, such as Llama. This innovative infrastructure not only enhances efficiency but also empowers data scientists to focus more on developing their models rather than managing the underlying technology.
Learn more about Amazon SageMaker HyperPod
Pricing
Free Version:
Yes
Company Details
Company:
Thinking Machines Lab
Headquarters:
United States
Website:
thinkingmachines.ai/tinker/
Media
Recommended Products
Ship Agents Faster
Gemini Enterprise Agent Platform lets you rapidly build, scale, govern and optimize production-ready agents grounded in your organization's data. The platform enables developers to build custom or pre-built agents for virtually any use case. New customers get $300 in free credits.
Product Details
Platforms
Web-Based
Types of Training
Training Docs
Customer Support
Online Support
Tinker Features and Options
Tinker User Reviews
Write a Review- Previous
- Next
