aid: efference-ai name: Efference AI description: Efference AI is a robotics perception and compute company building the hardware-plus-software foundation for data-driven robotics at scale. Founded in 2025 by Gianluca Bencomo and based in San Francisco (Y Combinator Fall 2025), Efference treats machine vision as a software problem, pairing affordable stereo-camera hardware with learned 3D-perception foundation models ("cheap hardware + great algorithms"). Its M1 device targets distributed, robotics-grade data collection at scale, while the H1 stereo system is built for real-time cloud inference with foundation models. The company serves autonomous vehicles, humanoid robots, and drones, and publishes hardware-interfacing documentation and firmware through its GitHub organization. As of this profile Efference exposes no public developer API, OpenAPI, or authenticated service surface. url: https://raw.githubusercontent.com/api-evangelist/efference-ai/refs/heads/main/apis.yml x-type: company x-source: vc-portfolio x-backed-by: - 500-global x-tier: stub x-tier-reason: portfolio-lead accessModel: pricing: unknown onboarding: unknown trial: false try_now: false public: false label: Unknown confidence: low source: [] generated: '2026-07-22' method: derived image: https://raw.githubusercontent.com/EfferenceAI/docs/main/logo/light.png specificationVersion: '0.20' created: '2026-07-17' modified: '2026-07-19' tags: - Company - Robotics - Computer Vision - Perception - Hardware - Stereo Cameras - Edge AI - Autonomous Systems - Foundation Models apis: [] maintainers: - FN: Kin Lane email: kin@apievangelist.com - FN: APIs.json email: info@apis.io common: - type: DomainSecurity url: security/efference-ai-domain-security.yml - type: Website url: https://efference.ai - type: Documentation url: https://docs.efference.ai - type: GitHubOrganization url: https://github.com/EfferenceAI - type: Support url: https://efference.ai/contact/ x-enrichment: date: '2026-07-19' status: backfilled pass: local-v1 note: backfilled from .gitignore signal + verified work evidence