# Deci AI

**Canonical:** https://apis.io/providers/deci-ai/  
**Website:** https://deci.ai/  
**APIs profiled:** 0

Deci AI was a deep-learning development company (founded 2019, Tel Aviv) whose platform helped teams build, optimize, and deploy computer-vision and NLP models in production. Its stack combined the AutoNAC neural-architecture-search engine, the Infery inference runtime, and the open-source SuperGradients PyTorch training library — the home of the YOLO-NAS object-detection architecture. Deci was acquired by NVIDIA in 2024; deci.ai now redirects to NVIDIA and there is no live standalone hosted API or developer console. This API Evangelist profile tracks the company's surviving first-party open-source software surface (SuperGradients and DataGradients on PyPI, published from the Deci-AI GitHub org).

## Kin Score — 6.8 / 100 (minimal)

Scored 2026-08-20 under rubric 0.12.0. Trend: flat (+0.0 from 6.8).

| Facet | Score |
|---|---|
| Discoverability | 50.0 |
| Contract Quality | 0.0 |
| Governance | 0.0 |
| Contract Governance | 0.0 |
| Operational Transparency | 2.6 |
| Developer Ergonomics | 7.1 |
| Commercial Clarity | 0.0 |
| Access Clarity | 0.0 |

## Agent readiness — 0.0 (human-only)

| Dimension | Value |
|---|---|
| Spec Presence | no |
| Agentic Access | no |
| Reversibility Documented | no |
| MCP Server | no |
| Auth Clarity | no |
| Idempotency | no |
| Error Semantics | no |
| OpenAPI Examples | no |
| Rate Limit Signal | no |
| Event Surface Described | no |
| Agent Skills | no |
| Well Known Catalog | no |
| Consent Identity | no |
| Agent Card | no |
| Dry Run Mode | no |

## Access

Unknown — onboarding: unknown, pricing: unknown, trial: no (confidence: low).

## Tags

Company, Artificial Intelligence, Machine-Learning, Deep Learning, Computer-Vision, Model Optimization, Inference, Neural Architecture Search, MLOps

---

Profiled by [API Evangelist](https://apievangelist.com) and published on [APIs.io](https://apis.io/providers/deci-ai/). Scores are computed from the provider's own public artifacts under a published rubric.