GPT Terra

OpenAI's mid-tier GPT-5.6 model, between Luna and Sol, available in ChatGPT and Codex.

HN commenters like Terra as an everyday model for coding and implementation1234, with some preferring its code and writing to Claude's567 and others citing lower cost and faster responses.89 The most common caveat is that many hand harder planning, debugging and review work to Sol or Astra10111213, and a few see it as stuck in an awkward middle tier.1415
The HN verdict, written from 15 cited comments
12recommend
1warn against
Jul 12, 2026first mention
Oct 26, 2026latest mention
Nov 25 – Oct 26
Used forfor codingfor subagent pipelinesfor data extractionfor cost-conscious usersfor Codex userscodingfeature additions with clear specs

Why people recommend it

Everyday coding · 3

Used as the default model for most implementation and coding tasks.

“If I can afford some time I use Terra (I use terra for most coding things)”

“Terra Medium/High for most implementations”

Show 1 more

“Terra xhigh is the only model I use for coding and it is good for literally everything.”

Multi-model workflows · 3

Commonly slotted into a role in multi-model setups, such as delegate, investigator or builder.

“Terra for delegate”

“Run Y copies of Terra in parallel to investigate the hypothesis results”

Show 1 more

“I use Opus to plan and GPT 5.6 Terra to build”

Versus Claude · 3

Some find it one-shots decent code more often than Sonnet and writes more concisely than Claude.

“Terra has been absolutely amazing for me and totally blows Sonnet out of the water.”

“GPT-5.6 Terra has been doing a significantly better job at writing concise PR bodies and commit messages than Claude's usual multi paragraph word soup.”

Show 1 more
o_mSep 12, 2026recommends

“Now I use GPT 5.6 Terra high fast mode, with Luna for everything else.”

Cost and speed · 3

One production user halved costs and cut response times by 50% by switching from GPT-5.5, and others call it faster than Sol.

“The bulk of our workload is now going via terra, which has cut our cost in half by itself, as well as improved response times by 50%”

runs production workflows extracting structured dataswitched from GPT-5.5on “Migrating a production AI agent to GPT-5.6: 2.2x fas…”

“Terra xhigh is the only model I use for coding and it is good for literally everything.”

Show 1 more

“I used to run it as my main driver at medium/high”

Caveats and dissent

Hard tasks · 3

Many users switch to Sol or Astra for hard bugs, planning, and complex reviews.

“Terra Medium/High for most implementations”

Show 1 more

Middle-tier squeeze · 2

Some see it as caught between cheaper Luna and stronger Sol, and one user moved to Luna xhigh.

“Terra have turned into this middle child in a weird spot”

“I used to run it as my main driver at medium/high”

Usage limits · 2

One user switched to it from Sol to manage limits and still doubts that will be enough; another saw a brief capacity notice.

“I've so far switched from Sol to Terra but not sure even that will be sufficient.”

“I love Terra and use it all of the time.”

Account warnings · 1

One user got a Cyber Abuse warning on their account after asking Astra and Terra about spoofing data.

“I started asking Astra and Terra about this”

Often compared with

Who recommends it in their own words

All comments 22

“I'm guessing Terra is good for the tasks that you'd typically use Sol 6.1-low for”

“and sometimes Terra level models”

“If I can afford some time I use Terra (I use terra for most coding things)”

“I used to run it as my main driver at medium/high”

“Terra, Luna - for implementation in simple routine cases.”

“Terra for delegate”

“Terra Medium/High for most implementations”

“Terra has been absolutely amazing for me and totally blows Sonnet out of the water.”

o_mSep 12, 2026recommends

“Now I use GPT 5.6 Terra high fast mode, with Luna for everything else.”

“I started asking Astra and Terra about this”

“Terra have turned into this middle child in a weird spot”

“GPT-5.6 Terra has been doing a significantly better job at writing concise PR bodies and commit messages than Claude's usual multi paragraph word soup.”

“I've so far switched from Sol to Terra but not sure even that will be sufficient.”

“I’ve been giving Terra first bite recently and it does pretty well”

“Terra xhigh is the only model I use for coding and it is good for literally everything.”

“I love Terra and use it all of the time.”

“Run Y copies of Terra in parallel to investigate the hypothesis results”

“The bulk of our workload is now going via terra, which has cut our cost in half by itself, as well as improved response times by 50%”

runs production workflows extracting structured dataswitched from GPT-5.5on “Migrating a production AI agent to GPT-5.6: 2.2x fas…”

“I use Opus to plan and GPT 5.6 Terra to build”

Threads where it came up

Also written as GPT-5.6 Terra, GPT 5.6 Terra. Data: gpt-terra.json. Quotes are verbatim; the verdict and theme labels are written by a model from these comments, and every clause links to its sources.

Keyboard shortcuts
j / kNext / previous item
o or EnterOpen selected item
/Search
g hHome
g tTop
g rRising
g cCategories
tToggle dark mode
?This help