Is AI Dumber Today? Community-tracked AI experience, updated hourly
← All recommendation scenarios

Current Top 3

Code generation and modification

A rank is relative: #1 can still have mixed or negative reputation.
#1

GPT-4o

GPT · 1 independent user

Observing

100% positive 0% mixed 0% negative

One language has too little evidence for a comparison.

100% positive 0% mixed 0% negative

Frequently praised

—

Frequently criticized

—

Original user comments

what ended up working for me was using gpt-4o for the planning phase, honestly the new one is pretty solid at breaking down a feature into clear steps and catching design issues before i even open claude code.

reddit · en

0% positive 0% mixed 0% negative

Frequently praised

—

Frequently criticized

—

Original user comments

No qualifying original comments in this slice.

100% positive 0% mixed 0% negative

Frequently praised

—

Frequently criticized

—

Original user comments

what ended up working for me was using gpt-4o for the planning phase, honestly the new one is pretty solid at breaking down a feature into clear steps and catching design issues before i even open claude code.

reddit · en
#2

GPT-5.6 Sol

GPT · 1 independent user

Observing

100% positive 0% mixed 0% negative

One language has too little evidence for a comparison.

100% positive 0% mixed 0% negative

Frequently praised

  • code correctness ×1

Frequently criticized

—

Original user comments

GPT 5.6 Sol over engineers just about everything.

hackernews · en

0% positive 0% mixed 0% negative

Frequently praised

—

Frequently criticized

—

Original user comments

No qualifying original comments in this slice.

100% positive 0% mixed 0% negative

Frequently praised

  • code correctness ×1

Frequently criticized

—

Original user comments

GPT 5.6 Sol over engineers just about everything.

hackernews · en
#3

Gemma 4 26B A4B

Gemma · 1 independent user

Observing

100% positive 0% mixed 0% negative

One language has too little evidence for a comparison.

100% positive 0% mixed 0% negative

Frequently praised

  • code correctness ×1

Frequently criticized

—

Original user comments

"I'd recommend Gemma 4 26B-A4B, with the caveat that it is best used as a coding assistant/general chatbot rather than a vibe coder - it will struggle with broad-scoped or vague prompts. Many deride Gemma for coding here, but with well-described, narrow prompts, it performs quite well in my experience."

reddit · en

0% positive 0% mixed 0% negative

Frequently praised

—

Frequently criticized

—

Original user comments

No qualifying original comments in this slice.

100% positive 0% mixed 0% negative

Frequently praised

  • code correctness ×1

Frequently criticized

—

Original user comments

"I'd recommend Gemma 4 26B-A4B, with the caveat that it is best used as a coding assistant/general chatbot rather than a vibe coder - it will struggle with broad-scoped or vague prompts. Many deride Gemma for coding here, but with well-described, narrow prompts, it performs quite well in my experience."

reddit · en