Use an uncensored model in Zed
Zed supports OpenAI-compatible model providers in its settings, so you can add Obscurify.ai and use an uncensored model in Zed’s assistant and agent. You get 200+ models through one key and an account number instead of an email.
Config verified against Zed’s language_models settings on 2026-11-08 (not run
here). Obscurify.ai’s OpenAI-compatible API itself is tested in our
OpenCode and Cline tutorials.
Before you start
- Zed, current release.
- An Obscurify.ai API key. Create an account —
a random 16-digit number, no email. Add credit on the
account page, then click Show API Key. The
key starts with
sk-obs1_and can be fetched again from your account number. New accounts may get $1 of starter credit while this month’s promo budget lasts.
Add the provider
Open Zed’s settings.json and add an openai_compatible provider under
language_models:
{
"language_models": {
"openai_compatible": {
"Obscurify": {
"api_url": "https://obscurify.ai/v1",
"available_models": [
{
"name": "pqhaz/deepseek-v4.1-flash-1017-heretic",
"display_name": "DeepSeek Heretic (Obscurify)",
"max_tokens": 128000,
"capabilities": { "tools": true }
}
]
}
}
}
}
Then open the assistant panel, pick the Obscurify provider, and paste your
sk-obs1_... key when Zed asks for it (keys are entered in the UI, not the file).
Pick a model
Any id from /v1/models in available_models,
keeping its prefix:
- Uncensored, needs Pro:
z-ai/glm-5.3-flash-uncensoredordeepseek/deepseek-v4.1-flash-uncensored. - Decensored, any funded account:
pqhaz/deepseek-v4.1-flash-1017-heretic.
Choose a model with "tools": true if you want Zed’s agent to edit files; some
Zed versions have had trouble with file edits on custom OpenAI providers, so test
a small edit first.
If something fails
- 402: no balance — add credit on the account page.
- 403
model_not_on_plan: uncensored models need Pro; decensored models work on any funded account. - 404
model_not_found: check the id against/v1/modelsand keep its prefix. - Agent won’t edit files: pick a tool-capable model and keep
"tools": true.
What stays private
Prompts and responses aren’t stored on Obscurify.ai’s servers; models with the ZDR tag run only on zero-data-retention routes. Providers see the prompt text but not who sent it. See the privacy policy.
More agents: Obscurify.ai for agents.