aquin
Aquin Labs develops AI systems for applied ML and data science, including Aleph for schema-bound prediction across tables and media. ML teams and researchers can also use its described developer environment and RL agent to build, test, and refine models.
Who it's for
- Applied ML teams working on prediction, annotation, or decision tasks across tables and media.
- Foundation-model researchers who need an environment for designing, testing, and refining models.
- Developers working in coding or research environments who want an RL agent that learns from harness lessons and refinement loops.
Not the right fit if…
- People seeking a conversational chat assistant: Aquin explicitly says Aleph is not a chat assistant and describes schema-bound outputs instead.
How it fits your workflow
For prediction work, Aleph is described as taking on tasks across tables, media, annotation, and decisions and returning schema-bound outputs such as probabilities, labels, boxes, or scores, with confidence information.
For model development, aq provides a developer environment with frameworks, libraries, tools, and agents for designing, testing, and refining foundation models. axiAuto is described as controlling developer environments through coding and research workflows, using lessons and refinement loops.
Pricing
The vendor doesn't publish prices. Check with aquin directly.
Prices checked on Nov 1, 2026 from the vendor's site. They can change; confirm before you buy.
Key features
- Aleph universal predictor
- Aquin describes Aleph as a predictor for applied ML and data science across tables, media, annotation, and decision tasks, with models intended to keep learning as work arrives.
- Schema-bound outputs
- Aleph is described as producing outputs such as probabilities, labels, boxes, and scores, along with confidence information for identifying uncertain results.
- aq developer environment
- A developer environment with frameworks, libraries, tools, and agents for designing, testing, and refining foundation models.
- axiAuto RL agent
- A self-learning agent for controlling developer environments, using harness lessons, rejected paths, and refinement loops.
- Model interpretability research
- Aquin lists work involving transformers and LLMs, attribution, sparse autoencoders, simulation, training observation, embedding models, evaluations, security, and benchmarks.
Limitations to know
- Aquin explicitly distinguishes Aleph from a chat assistant; the page describes schema-bound prediction outputs rather than conversational assistance.
- The page does not provide pricing or an onboarding process.