Superconductor Blog

How we build, benchmark, and operate coding agents.

What is Multiplayer AI?

Multiplayer AI means people on the same team working with the same AI agents, in shared sessions and with shared learnings, instead of each person alone in a private chat.

What is Multiplayer AI?

Launching the Multiplayer AI Manifesto

Grok 4.6 joins the Pantheon

Kimi K3 is the best open-weight model on our custom SWE-Bench. It's also the slowest.

What is Multiplayer AI?

Weekly Changelog: August 30-September 5

Launching the Multiplayer AI Manifesto

Introducing Chats: talk to an agent about your project, without writing code

Multiplayer conversation updates: See who's in the session with you

Weekly Changelog: August 23-29

Weekly Changelog: August 16-22

Now available: Linear MCP integration

Weekly Changelog: August 9-15

Grok 4.6 joins the Pantheon

Weekly Changelog: August 2-8

Now available: Zapier MCP integration

Kimi K3 is the best open-weight model on our custom SWE-Bench. It's also the slowest.

Weekly Changelog: July 26-August 1

Why coding agents belong in remote sandboxes

Now available: Figma MCP integration

Weekly Changelog: July 19-25

Preventing agents from cheating on our custom SWE-Bench

Weekly Changelog: July 12-18

GPT-5.6 lost our coding benchmark. We switched to it anyway

Weekly Changelog: July 5-11

Weekly Changelog: June 28-July 4

Cursor agents in Superconductor

Factory Droid agents in Superconductor

Pi agents in Superconductor

Weekly Changelog: June 21-27

Weekly Changelog: June 14-20

Bring your Notion workspace into Superconductor

Automations and Recurring Tasks in Superconductor

Bring Granola meeting context into Superconductor

Weekly Changelog: June 7-13

Grok Build and Composer agents in Superconductor

Introducing: Superconductor desktop app

Weekly Changelog: May 31-June 6

Weekly Changelog: May 24-30

Teaching coding agents to debug Rails memory issues with derailed_benchmarks

Weekly Changelog: May 17-23

Now available: Atlassian Rovo MCP integration

Now available: Better Stack MCP integration

Grok Build is surprisingly competitive on our Personal SWE-Bench

Weekly Changelog: May 10-16

Weekly Changelog: May 3-9

QA Review: Check agents' work before you review the diff

Weekly Changelog: Apr 26 - May 3

Weekly Changelog: Apr 19-25

GPT-5.5 is out, and it's the new quality-per-dollar leader

Weekly Changelog: Apr 12-18

Implementation Recommendations: Let AI pick the best implementation

Weekly Changelog: Apr 5-11

Weekly Changelog: Mar 29 - Apr 4

Weekly Changelog: Mar 23-29

Weekly Changelog: Mar 16-22

Now available: PostHog MCP integration

We tested 6 cloud agent platforms on the same task: Here's what happened

Recent Activity sidebar: Keep track of your implementations

Weekly Changelog: Mar 8-14

Weekly Changelog: Mar 1-7

Automatic Environment Setup: AI configures your cloud dev env

Now available: Filter diffs to see uncommitted changes

Introducing Meeting Assistant: Real-time ticket creation from your meetings

Live preview pane now shows sandbox startup commands

Network sandboxing: Set network access rules for coding agents

GPT-5.3 Codex vs Opus 4.6: We benchmarked both on our production Rails codebase

Now available: Guided review

Support for OpenAI premium plans

Support for PRs created outside of Superconductor

Star your favorite projects, tickets, and implementations

Generate and understand videos (and more) with Agent Skills

Workspaces: Transforming how teams work with AI

Ticket Artifacts: See agents' outputs right on the ticket

Agent Benchmarks

Now available: Sentry MCP integration

Now available: iOS app

Now available: GitHub integration

Now available: OpenCode models

Now available: Slack integration