Is /dev/fast profitable or generating revenue in 2026?

No. As of July 2026, /dev/fast has no disclosed revenue, customers, or pricing. Its product is invite-only, the review tool is labeled alpha, and its trace-storage component remains private and unbuilt, per its own site and YC listing.

Could AI coding tools like Claude Code or Cursor make /dev/fast irrelevant?

Possibly. /dev/fast's review layer is a thin wrapper over the same frontier coding agents (Claude Code, Codex, Cursor) that are commoditizing AI code review industry-wide, and GitHub itself can add similar features natively without conflict, as of July 2026.

Who is /dev/fast's biggest competitor?

CodeRabbit and Greptile. CodeRabbit had raised $88M total ($550M valuation) with 8,000+ paying customers as of September 2025; Greptile raised $30M total ($180M valuation) reviewing 500M+ lines of code monthly for clients like Brex and Substack as of January 2026.

/dev/fast is an independent, privately held startup founded by Sid (CEO), Milan Bhandari (CTO), Ketan Agrawal (CSO), and Alex (CPO), backed via Y Combinator's Winter 2026 batch, with no other disclosed investors as of July 2026.

Has /dev/fast raised a Series A?

No public Series A has been announced for /dev/fast as of July 2026. The company is currently operating on YC's standard batch investment with no disclosed outside funding, unlike rivals CodeRabbit ($60M Series B, September 2025) and Greptile ($25M Series A, September 2025).

What makes /dev/fast different from GitHub?

/dev/fast frames itself as a full 'AI-native code forge' rebuilding hosting, review, and CI/CD for agent-authored code, including a VPC-hosted trace-storage system for compliance. As of July 2026 this trace-storage feature is still marked private and unbuilt.

YC Winter 2026 · United States

/dev/fast

/dev/fast is a YC W26 startup building an AI-native code forge for agent-authored code.

Losing its protection

Assessed 27 July 2026 · Evidence as of 27 July 2026

The structural protection is eroding.

On the recordSince 27 July 2026 · Checkable from within 6 months post-Demo Day · Every call on the record Checkpoint: Product graduation from alpha/invite-only to general availability, or a Series A announcement

The call

/dev/fast is pre-revenue and invite-only in a category where CodeRabbit and Greptile already have funding, customers, and shipped products.

Dev/fast wants to be the tool that lets AI agents review other AI agents' code, positioning itself as a modern version of GitHub built for this new world. The problem is that bigger competitors like CodeRabbit and Greptile already have paying customers and serious funding, GitHub (which owns the original code-review market) could add these features itself without breaking its business, and AI tools like Claude and Cursor are already doing some of this work for free. The one thing that might save dev/fast is if it can ship a VPC-hosted system that lets companies audit and trace what their AI agents did, for security and compliance reasons, before competitors like Ellipsis build the same thing.

Why

AI pressure: Claude Code, Codex, Cursor and frontier-lab self-review tools. As AI models get better at reviewing their own code, the whole review layer /dev/fast wants to own could just get absorbed into the coding agents themselves. Could AI do the core job? Yes, likely.

Bigger players: GitHub (Microsoft). GitHub has no reason not to copy AI-agent review and trace features natively once they matter, and it owns the developer base already. Could a giant copy it? Yes, easily.

What protects it. None shipped yet. VPC-hosted agent trace-storage for compliance is the one gap rivals haven't closed, but it's still private and in development.

What would change this call

/dev/fast announces a Series A or GA launch within 6 months of Demo Day. Shows it can convert alpha status into a fundable, sellable product.
Ellipsis or Replicas ships VPC-hosted agent trace-storage before /dev/fast. Would remove /dev/fast's only unbuilt structural differentiator.
GitHub ships native large-diff AI review or agent orchestration features. Shows the incumbent can absorb /dev/fast's positioning without a business-model conflict.

Key numbers

CodeRabbit: $88M raised, $550M valuation (2025-09-15)
CodeRabbit: 8,000+ paying customers (2025-09-15)
Greptile: $30M raised, $180M valuation (n.d.)

Named risks

No shipped moat; core review layer is a thin wrapper over commoditizing frontier coding agents. Severity: high
GitHub/Microsoft faces no conflict copying features natively, could absorb /dev/fast's wedge. Severity: high
Well-funded rivals CodeRabbit and Greptile already have thousands of customers in same niche. Severity: high
Only defensible feature, VPC-hosted trace-storage, remains unbuilt and is being raced by Ellipsis. Severity: medium
No disclosed funding beyond YC batch investment, no job postings, signaling fragile runway. Severity: medium

Rivals, as read in July 2026

Rival	Direction	Evidence
CodeRabbit	losing to	8,000+ customers, $550M valuation vs /dev/fast's zero disclosed customers, as of Sept 2025
Greptile	losing to	500M+ lines reviewed monthly, named enterprise logos (Brex, Substack) as of Sept 2025
Ellipsis	contesting	Ships VPC-adjacent sandboxed agent orchestration overlapping /dev/fast's unbuilt trace-storage, June 2026
GitHub (Microsoft)	losing to	Shipped native Copilot coding agent May 2025, can absorb review/orchestration features at will
Replicas	contesting	Used by 20+ YC startups for agent orchestration, same wedge as /dev/fast's multi-agent factory claim

Questions people ask

Is /dev/fast profitable or generating revenue in 2026?: No. As of July 2026, /dev/fast has no disclosed revenue, customers, or pricing. Its product is invite-only, the review tool is labeled alpha, and its trace-storage component remains private and unbuilt, per its own site and YC listing.
Could AI coding tools like Claude Code or Cursor make /dev/fast irrelevant?: Possibly. /dev/fast's review layer is a thin wrapper over the same frontier coding agents (Claude Code, Codex, Cursor) that are commoditizing AI code review industry-wide, and GitHub itself can add similar features natively without conflict, as of July 2026.
Who is /dev/fast's biggest competitor?: CodeRabbit and Greptile. CodeRabbit had raised $88M total ($550M valuation) with 8,000+ paying customers as of September 2025; Greptile raised $30M total ($180M valuation) reviewing 500M+ lines of code monthly for clients like Brex and Substack as of January 2026.
Who owns /dev/fast?: /dev/fast is an independent, privately held startup founded by Sid (CEO), Milan Bhandari (CTO), Ketan Agrawal (CSO), and Alex (CPO), backed via Y Combinator's Winter 2026 batch, with no other disclosed investors as of July 2026.
Has /dev/fast raised a Series A?: No public Series A has been announced for /dev/fast as of July 2026. The company is currently operating on YC's standard batch investment with no disclosed outside funding, unlike rivals CodeRabbit ($60M Series B, September 2025) and Greptile ($25M Series A, September 2025).
What makes /dev/fast different from GitHub?: /dev/fast frames itself as a full 'AI-native code forge' rebuilding hosting, review, and CI/CD for agent-authored code, including a VPC-hosted trace-storage system for compliance. As of July 2026 this trace-storage feature is still marked private and unbuilt.

Timeline

2026-07-21Third-party directory confirms /dev/fast is YC W26, invite-only, no public API/docs
2026-07-17YC developer-tools batch listing shows /dev/fast among small, early-stage W26 peers
2026-04Sacra estimates CodeRabbit ARR reached roughly $40M, up from ~$5M a year prior
2026-03Greptile ships v4 with base-plus-usage pricing, citing rising AI-agent commit volume
2026-02-03Coverage confirms CodeRabbit's prior $16M Series A (2024) ahead of its Series B
2026-01-05Reporting confirms Greptile at ~$30M total funding, ~$180M valuation
2025-09-23Greptile raises $25M Series A led by Benchmark and launches v3
2025-09-15CodeRabbit closes $60M Series B at ~$550M valuation, ships agent-output CLI reviewer
2025-05-19GitHub introduces Coding Agent for GitHub Copilot
2025/dev/fast founded in San Francisco, later joins YC Winter 2026 batch

Revision history

27 July 2026

First assessed. Losing its protection. The text above is fixed as of this date and is not edited afterwards.

In the same band

CellType OctaPulse The Token Company