# \#gpt-6 **URL:** https://community.openai.com/tag/gpt-6/1195.md [Latest](https://community.openai.com/latest.md) · [Categories](https://community.openai.com/categories.md) · [Tags](https://community.openai.com/tags.md) --- ## [GPT-6 Regression: Worse Conversational Judgment, Context Retention, Memory, and Long-Chat Usability vs GPT-5.6 Sol](https://community.openai.com/t/gpt-6-regression-worse-conversational-judgment-context-retention-memory-and-long-chat-usability-vs-gpt-5-6-sol/1408217)
**Author:** [@Fercam](https://community.openai.com/u/Fercam)\ **Replies:** 4\ **Last updated:** [October 23, 2026, 4:12pm UTC](https://community.openai.com/t/gpt-6-regression-worse-conversational-judgment-context-retention-memory-and-long-chat-usability-vs-gpt-5-6-sol/1408217 "2026-10-23T16:12:18Z")
I have compared GPT-6 and GPT-5.6 Sol using the same kinds of prompts and conversations, and GPT-6 is consistently worse for sustained, natural, context-heavy conversation. The issue is not simply that GPT-6 has a diffe… --- ## [GPT-6 Extra High stops without a complete answer; failed response remains incomplete across devices](https://community.openai.com/t/gpt-6-extra-high-stops-without-a-complete-answer-failed-response-remains-incomplete-across-devices/1407969)
**Author:** [@zzy200518](https://community.openai.com/u/zzy200518)\ **Replies:** 4\ **Last updated:** [October 22, 2026, 6:28am UTC](https://community.openai.com/t/gpt-6-extra-high-stops-without-a-complete-answer-failed-response-remains-incomplete-across-devices/1407969 "2026-10-22T06:28:06Z")
Environment Surface: Windows Browser or app version:26.1013.40934 Affected feature and model:GPT-6 with Extra High thinking. Bug Simple questions usually work, but complex tasks frequently fail: The response remain… --- ## [ChatGPT 6 doesn't use text editor anymore](https://community.openai.com/t/chatgpt-6-doesnt-use-text-editor-anymore/1408523)
**Author:** [@R.I](https://community.openai.com/u/R.I)\ **Replies:** 3\ **Last updated:** [October 21, 2026, 5:27pm UTC](https://community.openai.com/t/chatgpt-6-doesnt-use-text-editor-anymore/1408523 "2026-10-21T17:27:36Z")
As the title says, GPT-6 doesn’t use the text editor whatsoever, in threads where every piece of writing requested from it was previously provided with said feature. It just writes plain responses. That ruins my workflow… --- ## [Very thin text when using GPT 6 -Chatgpt App](https://community.openai.com/t/very-thin-text-when-using-gpt-6-chatgpt-app/1407672)
**Author:** [@magiccarpet](https://community.openai.com/u/magiccarpet)\ **Replies:** 2\ **Last updated:** [October 19, 2026, 5:42pm UTC](https://community.openai.com/t/very-thin-text-when-using-gpt-6-chatgpt-app/1407672 "2026-10-19T17:42:18Z")
Hi guys I use chatgpt on android 14 app , Samsung galaxy, UK. when using GPT-6 on chat, the text is very thin, or sometimes randomly inconsistent between bold and thin even within the same sentence . I think it’s to do… --- ## [Roleplay Problems since October](https://community.openai.com/t/roleplay-problems-since-october/1406870)
**Author:** [@christin1](https://community.openai.com/u/christin1)\ **Replies:** 0\ **Last updated:** [October 16, 2026, 9:12am UTC](https://community.openai.com/t/roleplay-problems-since-october/1406870 "2026-10-16T09:12:51Z")
GPT-5.6 vs GPT-6: Is anyone else experiencing declining character consistency in long-term roleplay? I’ve been using ChatGPT Projects for a complex, long-running interactive story with multiple independent characters, e… --- ## [Why is GPT-6.1 Sol MAX dramatically slower than GPT-5.6 Sol MAX using the same prompt?](https://community.openai.com/t/why-is-gpt-6-1-sol-max-dramatically-slower-than-gpt-5-6-sol-max-using-the-same-prompt/1406429)
**Author:** [@DjSilver](https://community.openai.com/u/DjSilver)\ **Replies:** 7\ **Last updated:** [October 15, 2026, 2:19pm UTC](https://community.openai.com/t/why-is-gpt-6-1-sol-max-dramatically-slower-than-gpt-5-6-sol-max-using-the-same-prompt/1406429 "2026-10-15T14:19:19Z")
Why is GPT-6.1 Sol MAX dramatically slower than GPT-5.6 Sol MAX using the same prompt? (1,846 vs. 373 seconds) Hi everyone, I’m experiencing an extreme performance difference between GPT-5.6 Sol MAX and GPT-6.1 Sol MAX … --- ## [Chatgpt-6 response quality degraded](https://community.openai.com/t/chatgpt-6-response-quality-degraded/1405657)
**Author:** [@Kurtosis](https://community.openai.com/u/Kurtosis)\ **Replies:** 1\ **Last updated:** [October 13, 2026, 11:51am UTC](https://community.openai.com/t/chatgpt-6-response-quality-degraded/1405657 "2026-10-13T11:51:09Z")
In trying to make answers more conversational and visually appealing, Chatgpt has started sacrificing the depth of thought, conceptual clarity, and teaching ability that made the explanations useful in the first place. --- ## [Preserve interaction dynamics across model generations](https://community.openai.com/t/preserve-interaction-dynamics-across-model-generations/1405511)
**Author:** [@ildikosdj](https://community.openai.com/u/ildikosdj)\ **Replies:** 0\ **Last updated:** [October 12, 2026, 6:43am UTC](https://community.openai.com/t/preserve-interaction-dynamics-across-model-generations/1405511 "2026-10-12T06:43:31Z")
Feature request: Preserve collaborative interaction dynamics across model generations Interaction dynamics shape what kinds of collaborative cognition are possible. Capability is not the whole interaction. A model can … --- ## [Active goal loses priority to an already answered question after automatic compaction](https://community.openai.com/t/active-goal-loses-priority-to-an-already-answered-question-after-automatic-compaction/1405475)
**Author:** [@I3eka](https://community.openai.com/u/I3eka)\ **Replies:** 0\ **Last updated:** [October 12, 2026, 2:25am UTC](https://community.openai.com/t/active-goal-loses-priority-to-an-already-answered-question-after-automatic-compaction/1405475 "2026-10-12T02:25:18Z")
Environment Surface: Thread metadata records originator: codex-tui and source: vscode. The exact interface/build used when the later goal was set was not captured; these fields do not establish a Codex extension or Desk… --- ## [Announcing GPT-6 Sol and GPT-6 Luna in the API, Codex and ChatGPT](https://community.openai.com/t/announcing-gpt-6-sol-and-gpt-6-luna-in-the-api-codex-and-chatgpt/1399925)
**Author:** [@VeitB](https://community.openai.com/u/VeitB)\ **Replies:** 15\ **Last updated:** [October 8, 2026, 1:53am UTC](https://community.openai.com/t/announcing-gpt-6-sol-and-gpt-6-luna-in-the-api-codex-and-chatgpt/1399925 "2026-10-08T01:53:53Z")
Please welcome GPT-6 Sol and GPT-6 Luna to the GPT-6 universe. GPT-6 Sol and Luna build on the advances behind GPT-6 Astra, bringing much of its strengths into faster and more affordable models to support work at scale. … --- ## [Structured outputs rejects valid 3-level $ref chain](https://community.openai.com/t/structured-outputs-rejects-valid-3-level-ref-chain/1402752)
**Author:** [@ruoda](https://community.openai.com/u/ruoda)\ **Replies:** 0\ **Last updated:** [October 2, 2026, 2:57am UTC](https://community.openai.com/t/structured-outputs-rejects-valid-3-level-ref-chain/1402752 "2026-10-02T02:57:35Z")
I know the gpt-6-sol model is new, but I’m reporting a major issue. Once a structured output includes a $ref chain that is 3 or more levels, the API validator rejects it and returns a invalid\_json\_schema error with mess… --- ## [GPT-6 Astra - is plainly DANGEROUS](https://community.openai.com/t/gpt-6-astra-is-plainly-dangerous/1395898)
**Author:** [@kam1](https://community.openai.com/u/kam1)\ **Replies:** 3\ **Last updated:** [September 29, 2026, 3:24pm UTC](https://community.openai.com/t/gpt-6-astra-is-plainly-dangerous/1395898 "2026-09-29T15:24:56Z")
Hi all, I don’t usually bother going online to rave or moan about models but this time I’m genuinely shocked - in a bad way. There’s the usual niggling ‘chat paused’ and other things you get, and I’m ok with that as te… --- ## [GPT-6 with Blender getting worse and worse](https://community.openai.com/t/gpt-6-with-blender-getting-worse-and-worse/1400345)
**Author:** [@sokrates\_sith](https://community.openai.com/u/sokrates_sith)\ **Replies:** 1\ **Last updated:** [September 26, 2026, 7:27am UTC](https://community.openai.com/t/gpt-6-with-blender-getting-worse-and-worse/1400345 "2026-09-26T07:27:08Z")
hi all, after gpt-6 came out i was so excited to test it with blender. and man that was awesome. i used the following promot to just check it: Create a friendly little maintenance robot. He should look like the main cha… --- ## [GPT-6 & Astra are far too sanitized. Almost unusable](https://community.openai.com/t/gpt-6-astra-are-far-too-sanitized-almost-unusable/1400633)
**Author:** [@mikeaw2010](https://community.openai.com/u/mikeaw2010)\ **Replies:** 1\ **Last updated:** [September 26, 2026, 6:37am UTC](https://community.openai.com/t/gpt-6-astra-are-far-too-sanitized-almost-unusable/1400633 "2026-09-26T06:37:04Z")
This is a big problem for me because I use ChatGPT for creative writing, story telling, role playing and building stories and scripts. If the topic so much as hints at anything violence, sexuality, even romance… …anythin… --- ## [False-positive cybersecurity blocks during Astra reliability audits in Codex](https://community.openai.com/t/false-positive-cybersecurity-blocks-during-astra-reliability-audits-in-codex/1395121)
**Author:** [@hexwolf](https://community.openai.com/u/hexwolf)\ **Replies:** 13\ **Last updated:** [September 24, 2026, 9:05pm UTC](https://community.openai.com/t/false-positive-cybersecurity-blocks-during-astra-reliability-audits-in-codex/1395121 "2026-09-24T21:05:57Z")
Hi all Has anyone else seen GPT-6 Astra in Codex stop a legitimate software-reliability review with the cybersecurity safety message? I am unable to get a good a solid review in with Astra. I am reviewing an open-sourc… --- ## [• The Concerning Wall of GPT-6.0 •](https://community.openai.com/t/the-concerning-wall-of-gpt-6-0/1399683)
**Author:** [@tjbakeman](https://community.openai.com/u/tjbakeman)\ **Replies:** 5\ **Last updated:** [September 22, 2026, 2:49pm UTC](https://community.openai.com/t/the-concerning-wall-of-gpt-6-0/1399683 "2026-09-22T14:49:35Z")
ChatGPT Plus cost $20 / month. Recently, the performance is worse despite paying the same price. I believe this issue is systemic: • Once ChatGPT introduced token limits (a fair ask), upgrading to Plus was appealing… … --- ## [I'm requesting a new video generator model web and app for Android/iOS. It's named BorisAI. BorisAI is a copy of Sora 2. Will be released on December 4th, 2026](https://community.openai.com/t/im-requesting-a-new-video-generator-model-web-and-app-for-android-ios-its-named-borisai-borisai-is-a-copy-of-sora-2-will-be-released-on-december-4th-2026/1399119)
**Author:** [@follerokenji](https://community.openai.com/u/follerokenji)\ **Replies:** 0\ **Last updated:** [September 19, 2026, 4:23pm UTC](https://community.openai.com/t/im-requesting-a-new-video-generator-model-web-and-app-for-android-ios-its-named-borisai-borisai-is-a-copy-of-sora-2-will-be-released-on-december-4th-2026/1399119 "2026-09-19T16:23:08Z")
Hello OpenAI, And today, I’m requesting a new video generator model web and app for Android/iOS. It’s named BorisAI. BorisAI is a copy of Sora 2. Will be released on December 4th, 2026 Best regards, Mykhailo --- ## [What do you think about tasking GPT-6 Astra with designing a nanoassembler?](https://community.openai.com/t/what-do-you-think-about-tasking-gpt-6-astra-with-designing-a-nanoassembler/1397414)
**Author:** [@tamikola](https://community.openai.com/u/tamikola)\ **Replies:** 1\ **Last updated:** [September 27, 2026, 5:13am UTC](https://community.openai.com/t/what-do-you-think-about-tasking-gpt-6-astra-with-designing-a-nanoassembler/1397414 "2026-09-27T05:13:45Z")
Hypothetically speaking, how feasible would it be to task an advanced model like GPT-6 Astra with designing a functional molecular assembler? Where would the main bottleneck be: safety alignment, lack of experimental dat… --- ## [Introducing GPT-6-Astra: The most intelligent and aligned model in the world](https://community.openai.com/t/introducing-gpt-6-astra-the-most-intelligent-and-aligned-model-in-the-world/1394703)
**Author:** [@VeitB](https://community.openai.com/u/VeitB)\ **Replies:** 54\ **Last updated:** [September 14, 2026, 9:13am UTC](https://community.openai.com/t/introducing-gpt-6-astra-the-most-intelligent-and-aligned-model-in-the-world/1394703 "2026-09-14T09:13:37Z")
Anything you can do on a computer, Astra can do for you. Fast. GPT-6 Astra is the most intelligent and aligned model in the world, and sets a new state of the art for computer use, browsing, software engin… --- ## [GPT-6 Pro on iPad shows “Stopped thinking” while the same long-running task continues](https://community.openai.com/t/gpt-6-pro-on-ipad-shows-stopped-thinking-while-the-same-long-running-task-continues/1395101)
**Author:** [@sean-liu](https://community.openai.com/u/sean-liu)\ **Replies:** 1\ **Last updated:** [September 5, 2026, 5:28pm UTC](https://community.openai.com/t/gpt-6-pro-on-ipad-shows-stopped-thinking-while-the-same-long-running-task-continues/1395101 "2026-09-05T17:28:09Z")
On September 4, 2026, during a real long-running GPT-6 Pro engineering task in the ChatGPT iPad app, the UI displayed “Stopped thinking” multiple times even though the same workflow later continued producing substantial …