Home›Companies›Reason Machines
Reason Machines

Reason is world's most effective agent on cost per task

Reason runs autonomous cloud agents with a frontier harness on cost per task. It delivers the same performance with 30-50% lower costs than other harnesses. By operating with a harness outside the sandbox, it acts seamlessly across cloud computers and local devices. Reason runs work from start to finish: from an issue to a finished pull request, on a schedule, or from a Slack alert. It is available as a Mac app, on the web, from the CLI, in Slack, and through a REST API, with about 100 integrations and shareable skills. Bring your own keys and subscriptions, and work rolls over to the next one when limits are hit.
Active Founders
Adi Singh
Adi Singh
Cofounder
co-founder @ Reason Machines EECS @ UC Berkeley, Cybernetics NTNU Prev built a rust based IDE from scratch 21
Sven Myhre
Sven Myhre
Co-founder
Co-founder of Reason Machines (YC F26) - Building autonomous software-engineering agents. Studied EECS at UC Berkeley as a NORAM Scholar and Cybernetics & Robotics at NTNU. Won the 2025 Norwegian AI Championship and “Most Technically Impressive” at the Supabase AI Hackathon at YC, built stochastic prediction models at F100, and was president of Cogito, Norway’s largest learned society for AI.
Company Launches
Reason: the most efficient agent harness
See original launch post

Hey everyone, we’re Adi and Sven, co-founders of Reason Machines 👋

TL;DR We built the most efficient agent harness, so teams can maximize their AI usage at up to 30-80% lower cost per task than Codex, Cursor, or Devin. Sessions run on cloud VMs that move across your devices, use your existing keys and subscriptions, and work directly in Slack.

Since launching 4 weeks ago: 20,000 users, 14,000 X followers, and 3M+ impressions.

https://www.youtube.com/watch?v=LVRSTgnaexg


Problem

AI models are finally good enough for real work. The bottleneck is no longer capability—it’s cost: how much work you get per dollar of inference.

That cost is set by the harness more than the model: the same model can cost up to 5× more depending on the harness. Today, teams are stuck choosing between simple chat bots that can’t touch infrastructure, and closed cloud platforms that lock you into one vendor’s models and markup.


Solution

We designed a minimal harness for the most work per token, with the power of a cloud VM and the ease of a personal bot:

  • Cloud VMs that follow you: Runs across cloud Linux, cloud Mac, local machines, or SSH boxes without losing state.
  • Swarm-scale parallelism: Run hundreds of sessions in parallel to ship a feature, or let workflows run for days.
  • BYO keys & subscriptions: Plug in your API keys, open weights, and Grok, Copilot, or Kimi subscriptions with auto-failover.
  • Slack-native: Mention @Reason in a thread to get back a PR, a deck, or an answer.

The harness

A minimal harness (bash + async code execution) beats labs' bloated harnesses on cost per task without sacrificing capability. On the Coding Agent Index, Reason scored 65.1 vs Codex’s 62.5, at 22% lower cost per solved task. Read the full benchmark write-up.

Coding Agent Index against cost per task, with Reason highest at about $4 per task, beside a comparison of Reason finishing a task in 3 model calls and a call-per-tool harness in 7


Team

Adi and Sven studied EECS at UC Berkeley. Adi worked in applied ML at Scale AI and previously built a Rust-based IDE; Sven built AI models at a Fortune 100 company.

More AI usage, for less cost.

Adi and Sven, cofounders of Reason Machines


Our ask

  • Try it: start on the web or add @Reason to Slack. YC companies get $1,000 in credits via our Deal.
  • Intros: We’d love intros to teams hitting AI usage limits or paying heavy bills for Codex, Cursor, or Devin.
  • Investors: We’re raising our seed round—reach us at founders@reasonmachines.com.
Previous Launches
The IDE that thinks ahead of you
YC Photos
Jobs at Reason Machines
San Francisco, CA, US
0.50% - 2.00%
Any (new grads ok)
Reason Machines
Founded:2026
Batch:Fall 2026
Team Size:2
Status:
Active
Location:San Francisco
Primary Partner:Aaron Epstein