Know what is answering.

BrokenGPT is the product and API layer. The model answering your requests is broken-one, a mixture of open-source language models. We do not name the specific underlying models, architectures, or providers that make up the mix.

Current configuration

Public alias: broken-one
Display name: broken-one
Advertised context cap: 65,536 tokens
Runtime: operator-configured model

The public alias broken-one stays the same across infrastructure changes, so a stable identity is always available even as the underlying mixture evolves.

What “highly steerable” means

The product aims to reduce unnecessary refusals on lawful requests. It does not promise zero moderation or guaranteed compliance. Platform-level hard boundaries remain in force even when a model would otherwise answer.

Accuracy and limitations

Model output can be false, biased, incomplete, outdated, or insecure. Do not rely on it as the sole basis for medical, legal, financial, safety-critical, or high-impact decisions. Verify code and factual claims independently.

Data flow

Web-chat messages are stored in Postgres to provide history. Requests sent directly to /v1/chat/completions are not stored as conversation content by this application; usage metadata such as token counts, model, latency, key ID, and request status is recorded. The configured inference provider may have its own retention terms, which must be disclosed here before launch.

License and provenance

Operators must verify commercial rights for the exact base model and every derivative or quantization. Preserve required notices, pin immutable revisions, verify checksums, prefer safetensors, and do not enable unreviewed remote code for community uploads.