REVIEWS / AI CODING / KILO UPDATED OCT 3, 2026 · 474 SOURCES

THE PRODUCT

Kilo

Kilo

VS Code AI coding agent with loyal fans; its new extension migration drew rollback complaints and its Code Reviewer has repo-exploration limits.

AI CODING HIGH CONFIDENCE

THE VERDICT

4.8

REALITY SCORE · OUT OF 10 · CONFIDENCE HIGH

COMPOSED FROM

USERS 4.8 · 471 voices · 100%
CRITICS no published scores yet

SENTIMENT · 474 REVIEWS

+ 33% positive · 22% neutral − 45% negative

OUR VERDICT

WE DON'T RECOMMEND THIS
Score 4.8/10 — no affiliate link by editorial policy
See alternatives →

// Honest verdicts are the whole point. We only monetise products we'd actually recommend.

●23 YOUTUBE ●75 HN ●364 LEMMY ●6 STACK EXCHANGE ●3 PRODUCTHUNT
USER● n=474
VIDEO● n=3
BRAND● AVAILABLE
INTERNET● n=0

AT A GLANCE · QUOTABLE

  • Rating: 4.8 / 10 (high confidence)
  • User voices: 474 across 5 platforms
  • Sentiment: 33% positive · 45% negative
  • Updated: Oct 3, 2026

GYIBB rates the Kilo 4.8/10 based on 474 user voices from 5 platforms. Confidence: high. Source: https://gyibb.com/ai-coding/kilo

BUY IF

Devoted user base on its own channel: 'I adore my Kilo agent', 'Our Kilo is the best'

  • + Exceptional reach for channel size: launch video at ~1.16M views on 51.8K subs
  • + Active development: new VS Code extension plus a dedicated AI Code Reviewer product
  • + Clear competitive niche: benchmarked against Roo Code and Cline in 2026 reviews

SKIP IF

Extension migration broke workflows: custom mode instructions lost, no UI config for modes/MCPs

  • − User reports vector-database code indexing removed: 'You've taken a serious wrong turn!'
  • − Real-time code-change monitoring no longer default in the VS Code UI after update; users rolling back
  • − Code Reviewer can't explore the full repo — 'severely limited' per user comment

Where the layers disagree ⚡

5 CONTRADICTIONS DETECTED

USER layer is unusable: the 471 'comments' are HN threads about kilometers, kilobytes and a 1KB chess engine — a keyword collision with 'Kilo' — so no user-vs-video cross-check is possible.

VIDEO VS USER

VIDEO internal split: the same launch video's comment section holds both the strongest praise ('Our Kilo is the best') and the harshest criticism ('You've taken a serious wrong turn!') — affection for the agent vs. regression in the new extension.

VIDEO VS USER

VIDEO vs VIDEO: the vendor markets 'Kilo Code Reviewer | AI Powered Code Reviews' while a user comment states it 'cannot explore the rest of the repo', capping the headline feature.

BRAND VS VIDEO

VIDEO (unofficial) gap: the one independent comparison (Tinkr, Roo vs Kilo vs Cline) has 2,365 views and zero substantive verdicts — no external validation of the vendor-channel claims.

BRAND VS VIDEO

NO CROSS-LAYER CHECK POSSIBLE: INTERNET (expert reviews) and BRAND (official claims) are empty, so video comments can't be tested against third-party data or the maker's own positioning.

BRAND VS VIDEO

WHERE THEY AGREE +

+ Devoted user base on its own channel: 'I adore my Kilo agent', 'Our Kilo is the best'
+ Exceptional reach for channel size: launch video at ~1.16M views on 51.8K subs
+ Active development: new VS Code extension plus a dedicated AI Code Reviewer product
+ Clear competitive niche: benchmarked against Roo Code and Cline in 2026 reviews

WHERE THEY DON'T −

− Extension migration broke workflows: custom mode instructions lost, no UI config for modes/MCPs
− User reports vector-database code indexing removed: 'You've taken a serious wrong turn!'
− Real-time code-change monitoring no longer default in the VS Code UI after update; users rolling back
− Code Reviewer can't explore the full repo — 'severely limited' per user comment
− No independent review data; all substantive feedback sits under the vendor's own videos

Where the 474 sources came from

VIEW EVERY CITATION →
YOUTUBE
23
HN
75
LEMMY
364
STACK EXCHANGE
6
PRODUCTHUNT
3

The four realities of the Kilo

Most review sites collapse everything into one number. We keep the layers separate so you can see where reality bends.

01
USER
n=474 · 5 platforms

What actual buyers say

DATA CONTAMINATION — treat as zero product signal. The 471 scraped 'user comments' are top-voted HackerNews threads matched on the word 'kilo', not on the product: a mph→km/h mental-math thread ('30 mph => 3 => 6 => 12 => 24 => 48 km/h'), a miles/kilometers and Celsius/Fahrenheit conversion debate ('22℃ -> 44 - 4 + 32 -> 72℉'), number-base theory ('every number base is base 10 when you interpret the 10 in that base'), a Fibonacci miles↔km conversion trick ('5 -> 8 means 500 -> 800'), and reaction threads to a ~1KB ('kilo'byte) chess engine ('such a tiny amount of code plays so well', 'played at the level of an earnest 12 year old'). Not one comment mentions the VS Code extension, AI coding agents, Roo/Cline, pricing, or the vendor. Conclusion: genuine user sentiment about Kilo Code cannot be computed from this sample; deriving a rating from it would be fabrication. The only real user-experience evidence in this dataset lives in the VIDEO layer.
02
VIDEO
n=23 · YouTube

What reviewers showed on camera

The only substantive layer. Two videos come from Kilo's own channel (51,800 subs): 'Kilo's New VS Code Extension is Live!' (~1,160,684 views — remarkably high reach for the sub count) and 'Kilo Code Reviewer | AI Powered Code Reviews' (4,193 views). Launch-video comments split sharply. Loyal camp: 'I adore my Kilo. agent :)', 'Naah, my Kilo is the best. lol. Our Kilo is the best.', plus forward-looking enthusiasm ('can't wait until the claude code leak is integrated into kilo'). Migration backlash is louder and more specific: 'Why did you break a perfectly functioning program? Where did you put the code indexing via the vector database? You've taken a serious wrong turn!'; another user reports the new-extension migration didn't carry over custom instructions for modes, that the old version's UI-based configuration of modes and MCPs is gone ('now it has become worse... I've switched back to the old version. I need to continue my work instead of configuring everything almost from scratch'), and that real-time monitoring of code changes in the main VS Code UI is no longer on by default. On the Reviewer video: a non-developer asks whether it edits code or only posts suggestions as comments, and a critical user notes 'since it cannot explore the rest of the repo, its reviewing abilities are severely limited unfortunately.' The sole third-party test — Tinkr's 'Roo Code vs Kilo Code vs Cline Review' (141K subs, 2,365 views) — produced no substantive verdict in its comments. Net read: engagement and brand affection are real, but release-quality regressions in the newest extension are the dominant complaint. Caveat: nearly all feedback sits under the vendor's own videos, so self-selection bias is likely.

Kilo's New VS Code Extension is Live!

Kilo · 1,160,684 views

"[comment] Wonderful...can't wait until the claude code leak is integrated into kilo. [comment] Why did you break a perfectly functioning program? Where did you put the code indexing via the vector database? You've taken a serious wrong turn…"

Kilo Code Reviewer | AI Powered Code Reviews

Kilo · 4,193 views

"[comment] So it doesn't edit the code. It just adds suggestions as comments correct?? Sorry I'm not a developer and I'm trying to learn to build stuff [comment] I adore my Kilo. agent :) [comment] since it cannot explore the rest of the rep…"

Roo Code vs Kilo Code vs Cline Review: Which AI Coding Agent Is Actually Better? (2026)

Tinkr | Reviews & Guides · 2,365 views

"[comment] в переводе через yandex браузер вообще интересно звузит. 😄 [comment] Why would you teach AI to talk like that 😭…"

03
INTERNET
n=0 · review sites

What the press said

No aggregate ratings were found for this product during the last harvest.
04
BRAND
official source

What the brand says

no brand page found

The official brand page was not successfully scraped during the last harvest.
Visit Official Site →

SIMILAR IN THIS CATEGORY

See all →
OpenCompanion

OpenCompanion

10.0

✓ Claimed: one-screen dashboard for multiple AI CLI sessions

Replay QA Security Scan

10.0

✓ Targets a credibly described gap: AI agents ship functional endpoints without auth checks (IDOR, broken access control, injection)

Interactive Sessions

Interactive Sessions

10.0

✓ Both hands-on and autonomous modes under one governance layer

Kiro Crew

Kiro Crew

10.0

✓ Enhances Kiro CLI significantly

DATA SOURCES & AUDIT

23
YOUTUBE
75
HN
364
LEMMY
6
STACK EXCHANGE
3
PRODUCTHUNT
3
YOUTUBE VIDEOS

474 data points across 5 platforms, synthesized via GYIBB's Truth Engine and fact-checked against source data before publication.

CONFIDENCE: HIGH · ANALYSED: OCTOBER 3, 2026 AT 07:18 PM · PROMPT V1.0 · READ METHODOLOGY →

Was this review helpful?

Embed this review

Writing about Kilo? Add the GYIBB verdict — free, no account needed.

GYIBB rating for Kilo 
<a href="https://gyibb.com/ai-coding/kilo" target="_blank" rel="noopener">
  <img src="https://gyibb.com/badge/ai-coding/kilo.svg" alt="GYIBB rating for Kilo" width="220" height="56">
</a>
← Back to all reviews

Kilo

GYIBB SCORE: 4.8/10

See alternatives →