Persistent HTTP 500 Internal Server Error on Antigravity - "Internal error encountered" since Feb 18/19 2026

Google Antigravity
1

Olá equipe do Google Antigravity / Gemini,

Estou enfrentando um erro persistente de HTTP 500 Internal Server Error ao tentar usar o Antigravity ( IDE desktop ). O erro acontece imediatamente ao enviar qualquer prompt ou ao tentar executar um agente, desde ontem (18 de fevereiro de 2026). Já atualizei o Antigravity para a versão mais recente, mas o problema continua.

Detalhes do erro:

  • Trajectory ID: 6a33c8c8-d2b1-…

  • Error: HTTP 500 Internal Server Error

  • Mensagem completa:

    text

    {
  "error": {
    "code": 500,
    "message": "Internal error encountered.",
    "status": "INTERNAL"
  }
}

  • Sherlog / TraceID: 0xe62ed2ed58d46093

  • Headers do response (principais):

    text

    {
  "Alt-Svc": ["h3=\":443\"; ma=2592000,h3-29=\":443\"; ma=2592000"],
  "Content-Length": ["109"],
  "Content-Type": ["text/event-stream"],
  "Date": ["Thu, 19 Feb 2026 05:00:22 GMT"],
  "Server": ["ESF"],
  "Server-Timing": ["gfet4t7; dur=3516"],
  "Vary": ["Origin","X-Origin","Referer"],
  "X-Cloudaicompanion-Trace-Id": ["e62ed2ed58d46..."],
  "X-Content-Type-Options": ["nosniff"],
  "X-Frame-Options": ["SAMEORIGIN"],
  "X-Xss-Protection": ["0"]
}

O que já tentei (sem sucesso):

  • Atualizar o Antigravity para a versão mais recente

  • Reiniciar o app/IDE completamente

  • Limpar cache do app e do browser

  • Resetar onboarding (Ctrl + Shift + P → “Antigravity: Reset onboarding”)

  • Logout e login novamente

  • Trocar de modelo (testei Gemini, Claude, etc.)

  • Verificar se a conta tem Google AI Pro/Ultra ativo (estou logado com conta LoggedIn, mas com acesso normal ao Antigravity)

Ambiente:

  • Sistema operacional: Ubuntu 24.04

  • Versão do Antigravity: 1.107.0

    1504c8cc4b34dbfbb4a97ebe954b3da2b5634516

    x64

  • Localização aproximada: Maputo, Moçambique

  • Tipo de uso: IDE desktop]

  • Modelo que estava usando quando deu erro: Gemini 3 Pro

  • Acontece em todos os prompts/agentes.

Muitos usuários estão relatando o mesmo erro exato nos últimos dias (veja threads como “HTTP 500 Internal Server Erro”, “Agent terminated due to error”, “Antigravity broken: getting only ‘Agent execution terminated due to error.’”, etc., no fórum). Parece ser um problema no backend/server-side (Server: ESF, trace IDs parecidos, text/event-stream), possivelmente outage ou instabilidade após updates recentes.

Podem confirmar se é um problema conhecido/generalizado e se há ETA para correção? Obrigado pela atenção e pelo ótimo trabalho no Antigravity!

Atenciosamente,
Zacarias (Maputo, MZ)

2

same situation here.

3

same HTTP 500 Internal Server Error

4

Thank you for bringing this to our attention. We have escalated the issue to our internal teams for a thorough investigation.

To ensure our engineering team can investigate and resolve these issues effectively, we highly recommend filing bug reports directly through the Antigravity in-app feedback tool. You can do this by navigating to the top-right corner of the interface, clicking the Feedback icon, and selecting Report Issue.

5

Opções:

  1. remova o cache local (no linux seria removendo ~/.config/Antigravity/)
  2. desinstale e instale novamente
  3. verifique se você está com a validação de 18anos implementada recentemente pelo google: Update your account to meet age requirements - Google Account Help
    ou
    https://myaccount.google.com/age-verification
6

Still broken as of May 7, 2026 — nearly 3 MONTHS of silence. Unacceptable.

I’m yet another developer adding to this thread because apparently that’s the only way to get Google’s attention. Persistent HTTP 500 Internal Server Error, TraceID: 0x67adb62783084100, Server-Timing showing dur=45623 — meaning your backend is timing out after 45+ seconds before throwing a generic “Internal error encountered.” Same error. Same silence. Different day.

Let’s be crystal clear about what’s happening here:

This issue was first reported on February 18/19, 2026. It is now May 2026. That is nearly 3 months of a broken product with ZERO official fix, ZERO ETA, and ZERO compensation for affected users — including paid subscribers.

The backend headers tell the full story:

  • X-Cloudaicompanion-Trace-Id confirms this is a Google infrastructure failure, not a user-side issue
  • SERVER-TIMING: dur=45623 = your servers are hanging for 45+ seconds before crashing
  • MODEL_CAPACITY_EXHAUSTED is the real underlying error being masked by the 500

Developers are losing real working hours every single day. We’re not hobbyists playing around — many of us have production workflows, automation pipelines, clients, and deadlines that depend on Antigravity being functional. Every day this stays broken is a direct cost to us.

What we are demanding — not asking, DEMANDING:

  1. An OFFICIAL acknowledgment from the Google Antigravity/Gemini team that this is a known, ongoing infrastructure issue
  2. A REALISTIC and SPECIFIC ETA for the fix — not “we are investigating”, not “thanks for the report”, an actual date
  3. COMPENSATION or quota extension for all paid users who have been affected during this downtime period
  4. A public status page update so developers stop wasting time debugging their own code for a problem that is 100% on Google’s end
  5. A post-mortem explaining how a critical service was left broken for 3+ months with no resolution

Google, you launched Antigravity aggressively, marketed it to developers, and then left us stranded with a broken tool and radio silence. This is not how you treat a developer community. This is not acceptable for a company of your scale and resources.

The community deserves better. Please escalate this internally — now.

7

I am also facing the same issue consistently.

8

Same issue, while gemini cli work normally

9

Hi @ankit_dev @Muhammad_Husnain1 @Zengming,

Could you please clarify which specific model you are utilizing when this issue occurs?

As a temporary workaround, you may find that reducing the input context, switching to an alternative model from the model selector or allowing a brief pause before retrying the request helps to resolve the issue.

10

Hahahaha, you’re lucky – you’ve only been experiencing this since February. I have accounts that have been experiencing this error since December 2025.

Google developers NEVER help with server-side issues – they just curse certain accounts.

Alternatives include Claude/Codex, which I’ve been using since early spring (because they don’t have 500 errors, models are accessible, and quotas are larger – they don’t get cut to 21 days like I had with one account instead of 5 hours).