Flash

Ai is Hot I guess

[Flash @GoogleDevss]

@OpenAI What fucking nonsense My account reset is 1 month duration before it used to take 1 weak now 1 month seriously?

[Flash @GoogleDevss]

@antigravity Hit a strange quota issue today. The dashboard showed ~60% quota remaining, but the platform still returned: "Baseline model quota reached" and blocked execution. Looks like a mismatch between displayed quota and actual available quota. Anyone else seeing this?

[Flash @GoogleDevss]

@antigravity They are making more limited usage for Claude models they run out of tokens within 8mins of running even through I have Google Pro its fucking useless just to run for 1 prompt atleast before they Update the Usage was still acceptable But you guys keep making things worst

[Flash @GoogleDevss]

@vikaskansalHQ @antigravity The daily quota was exhausted within 15 minutes, and then a single prompt consumed another 500 credits from my 1000-credit plan. That level of usage feels excessive and too restrictive for Pro users

[Flash @GoogleDevss]

@antigravity Previous Gemini Flash 3 models had far better limits because they were unlimited.
Now Google combined them with Gemini 3.1 Pro quotas, making usage very limited.
Separate quotas for each model would be better for all models Would be a Better solution and all Users Would bebetter

[Flash @GoogleDevss]

@antigravity It Would be better If you kept Flash models usage separate Just like Before and Flash 3.5 uses So many tokens at once, It used all my 1000 credits Plus Daily quota In a single Prompt But i was satisfied with the Output 👀

[Flash @GoogleDevss]

@antigravity Amazing it fixed my bugs and improved my image pipeline. When Claude Opus 4.6 hit its usage limit, I tried Gemini Flash 3.5 (High). It took 2–3 hours to fix everything and gave even better outputs.
Downside: it used all my 1000 credits and my daily quota In 1 prompt.

[Flash @GoogleDevss]

@antigravity I saw this Coming 👀 When it showed antigravity 2.0 installing

[Flash @GoogleDevss]

@antigravity Idk what fucking wrong with Gemini 3.1pro in antigravity it doesn't even know what tools and abilities it has until the user reminds it to use until then it fucking nonsense and also It keep thinking this useless stuff its clearly looping and not even follow the users instruction

[Flash @GoogleDevss]

@Samaytwt These are the two biggest areas that made antigravity Dead
Servers issues heavy traffic errors
Usage limits and cooldown timer
Users have been facing these problems for months, yet there has been little to no visible progress or updates That's why many users Stopped using it

[Flash @GoogleDevss]

@jahirsheikh8 There is Official announcement for this that antigravity is being Shutting down
There are alot of issues which users are facing like server issues heavy traffic error and usage limits cooldown

[Flash @GoogleDevss]

@antigravity Should focus on fixing Antigravity first. The main issues users are facing are:
Server issues
Usage limits
These are the two biggest areas that need improvement. Users have been facing these problems for months, yet there has been little to no visible progress or updates.

[Flash @GoogleDevss]

@Google Should focus on fixing Antigravity first. The main issues users are facing are:

Server issues

Usage limits

These are the two biggest areas that need improvement. Users have been facing these problems for months, yet there has been little to no visible progress or updates.

[Flash @GoogleDevss]

@sama Project S: A hybrid autonomous AI agent (Ollama + APIs). ⚙️
• Multi-task automation via Python
• Uncensored, 100% local image generation (RTX 4060 optimized)
• Browser task automation
• Cross-platform control (Windows & Android via Telegram Bot)
#AI #Ollama #AgenticAI

[Flash @GoogleDevss]

@OpenAI CodeX usage is getting ridiculous.
On GPT-5.4 Medium, limits are hit in just 1–2 prompts. Earlier it handled 4–5 prompts easily what changed?
Now it burns out instantly, and the cooldown takes nearly a full week. This makes it almost unusable for any serious work which feelswaste

[Flash @GoogleDevss]

@OpenAIDevs CodeX usage is getting ridiculous.
On GPT-5.4 Medium, limits are hit in just 1–2 prompts. Earlier it handled 4–5 prompts easily what changed?
Now it burns out instantly, and the cooldown takes nearly a full week. This makes it almost unusable for any serious work which feelswaste

[Flash @GoogleDevss]

@OpenAIDevs CodeX usage is getting ridiculous.
On GPT-5.4 Medium, limits are hit in just 1–2 prompts. Earlier it handled 4–5 prompts easily what changed?
Now it burns out instantly, and the cooldown takes nearly a full week. This makes it almost unusable for any serious work which feelswaste

[Flash @GoogleDevss]

@antigravity fix this madarchod
bullshit right now!
ALL models dead with HTTP 503

MODEL CAPACITY EXHAUSTED. Can't run

30 seconds.

Trajectory:

bc1b36d9-6168-4e4b-af61-11e0cb5d7583 You hyped this trash and it's completely broken. Scale your servers or fuck off

[Flash @GoogleDevss]

@antigravity @antigravity @antigravity fix this madarchod bullshit right now!
ALL models dead with HTTP 503 MODEL_CAPACITY_EXHAUSTED. Can't run 30 seconds.
Trajectory: bc1b36d9-6168-4e4b-af61-11e0cb5d7583
You hyped this trash and it's completely broken. Scale your servers or fuck off.

[Flash @GoogleDevss]

@OpenAI #OpenAIDevDay2026