Hey Everyone
Service is up and running. If API is not working please ping me.

Just checked. Deepseek Pro is throwing error. Deepseek Flash is working

Deepseek is up. Really sorry for downtime. We will assign more days. In the meantime we are working on better VPS and also a backup so this issue doesn't occur again. For now the all models are up and running 🔥

People who are facing issue in model hit me up.

We have created backup route if the main service goes down. They will be available if the maine service is unaccessible. We will also add service status so you can check whether service is yup or down.

Gpt and minimax is removed. We are settting up route and will be added back.

2 new model added.
Glm-5.3 and kimi-k3

z-ai/glm-5.3
moonshotai/kimi-k3

gpt-5.6-luna is also back

We would love to hear your feedback.

To use Jail Break Models use this
deepseek-v4-pro-jailbreak
anthropic/claude-opus-4.8-jailbreak
anthropic/claude-sonnet-5-jailbreak
deepseek/deepseek-v4-pro-0813-jailbreak
deepseek/deepseek-v4-flash-0731-jailbreak
openai/gpt-5.6-luna-jailbreak
z-ai/glm-5.2-jailbreak
qwen/qwen3.8-max-jailbreak
z-ai/glm-5.3-jailbreak
moonshotai/kimi-k3-jailbreak

🚨 JAILBREAK METHOD UPDATED

The jailbreak system has now been fully updated with our strongest method so far for 2026.

What changed:

better instruction persistence

stronger behavior consistency

improved handling for difficult prompts

more stable across supported models

This is a backend-side upgrade, so users don’t need to change anything.

⚠️ Jailbreak still does not mean zero refusal in every situation. Behavior can still vary by model and upstream provider.

More model updates are currently being tested and will be announced after verification.

Just to remind everyone 1 api is allowed for 1 person. If we see used by mutiple people or used as a gateway the api will be revoked without a refund. We have flagged few users. Final reminder for everyone.

Next Month Prcing will be updated. If you guys want to continue at same price let me know.

We are testing 0x Alpha and Opus 5, you can expect it soon.

We will be updating the token provider to api.hackwithclaude.com soon. So you can buy token for model available in unlimited as well. Including jail break.

There is some issue reported for 5.3 and 5.3 jailbreak we are working on it. In the meantime you can use other models like Opus and all they are working fine.

Tonight: GLM 5.3 Flash will be available on Hackwithclaude. ⚡
And there's another upgrade that we believe is even more important:
👁️ We will enable Vision support directly from our backend for models that natively support image input.
This means that Hackwithclaude will not only continue to add more model options, but we will also unlock the full potential of each model.
One API key. Many models. Coding, reasoning, automation, and now increasingly more multimodal workflows, all within a single gateway.
We want every update to make the key you already have feel even more valuable, rather than just adding another model name to the catalog.
🇬🇧 Tonight: GLM 5.3 Flash is coming to Hackwithclaude, and Vision support will be enabled from our backend for models that natively support image input.
More models. More capabilities. Same gateway.

New Model - z-ai/glm-5.3-flash is Live 🔥