Guide Labs blog · Jun 2026
Company announcement for Clarity, the first inherently interpretable AI platform, with concept explanations, training-data attribution, and concept steering.
Public company, workplace, funding, and market signals
Updated Jul 30, 2026
Guide Labs is a San Francisco-based AI research company building inherently interpretable foundation models and an interpretable AI platform so users can trace, steer, and audit model outputs and training-data provenance.
Primary product
Clarity, an inherently interpretable AI platform powered by Steerling-8B
Founded
2023
Headquarters
San Francisco, California, United States
Team size
10-20
Industry
Research Services
Sub-industry
Interpretable AI / foundation models
Offices
0 jobs at Guide Labs
Check back later for new openings
Business model
Leadership
Co-Founder & CEO
Co-Founder & CSO
Head of Operations and Business Development
Co-Founder
Founding Research Scientist
Investors
Mission-driven and research-heavy, with a strong emphasis on interpretability, reliability, and impact over formal credentials; the company explicitly encourages applicants who can demonstrate ability even without conventional background signals.
Compensation
Public job postings show pay transparency for some roles, including a Research Scientist Intern at $57/hour and a Research Engineer at $150k-$200k; broader compensation policy was not publicly disclosed.
Pricing
API / research preview; no public pricing disclosed
Differentiators
Technology
Customers
Estimated revenue
Third-party estimates vary widely, from roughly $440K ARR (GetLatka) to about $2.8M ARR (StartupHub); no company-filed revenue disclosure found.
Estimated monthly visits
6K
Traffic estimate as of Jun 2026
Guide Labs blog · Jun 2026
Company announcement for Clarity, the first inherently interpretable AI platform, with concept explanations, training-data attribution, and concept steering.
Guide Labs blog · Mar 2026
Explains inference-time concept suppression and steering for safety/alignment, including reported reduction of harmful outputs from 80% to 29% without finetuning.
Guide Labs blog · Mar 2026
Company essay framing its core thesis: models should be auditable, steerable, and understandable by design, with Steerling-8B as the proof point.
TechCrunch · Feb 2026
TechCrunch covered the launch of Steerling-8B, describing Guide Labs' approach to making every token traceable back to training data and positioning the company as a new entrant in interpretable LLMs.
FinSMEs · Dec 2025
Funding coverage describing a seed round led by Initialized Capital with participation from Tectonic Ventures, Y Combinator, Lombardstreet Ventures, E14 Fund, Pioneer Fund, and angels.