OneTriangle
Inference optimized from KV cache transfer
San Francisco, United States · Founded 2026
- Headquarters
- San Francisco, California
- Employees
- 1–10
- Founded
- 2026
- Business Model
- B2B
- Website
- onetriangle.ai
About
OneTriangle operates "the cheapest, fastest inference cloud," using frontier research on KV cache transfer between models to cut inference costs and time by 20%, running prefill on a small model and decode on a large one. The company offers "no API changes" integration and is "upstreaming it into vLLM," delivering "large-model quality at small-model cost" for teams running open-weight models.
Summary
OneTriangle is an Artificial Intelligence company based in San Francisco, United States, founded in 2026.
Tech & App Stack
2 technologies & apps tracked
Every product and vendor across these categories, with where each signal came from.
Create a free account with 25 free credits every month, no card needed, to unlock the full tech & app stack, funding history, and valuations.
Similar Companies
AI startup building artificial general engineer for physical-world robotics and automation.
Mistral co-creates tailored AI systems for enterprises and governments, combining frontier models...
Chinese AI company developing GLM large language models and multimodal agent APIs for enterprises...
Cursor is the best way to code with AI, and built to make you extraordinarily productive.
Provides AI inference hardware and cloud platform for enterprises, using RDUs and SambaCloud to r...
UiPath provides an enterprise automation platform that coordinates AI agents, software robots, pe...
Your company? Embed this profile
Add the live badge to your site — it stays current as your funding data updates.
<a href="https://indexed.vc/companies/onetriangle"><img src="https://indexed.vc/api/badge/company/onetriangle" alt="OneTriangle on Indexed.vc" width="260" height="76" /></a>