The KOLDOS API
- Local API
- Available
- Hosted API
- Planned
- Hosted access
- Manual approval
Getting started
From zero to a first turn
- 01
Start the local API
Run
Local API referencekoldos serveon the computer where KOLDOS is installed. It listens on 127.0.0.1:8128. - 02
Send a turn
Use any client that speaks the common chat completions format, with the model koldos. Build one below.
- 03
Ask about hosted access
The hosted API is planned. Planned window: Dec 30, 2026 to Feb 10, 2027. A planned window, not a fixed release date. Developer access is provided manually. Contact itzkoldkir on Discord to request API access.
Discord username
itzkoldkir
- 04
Keys and credits
When the hosted service starts, keys, usage and credits live in the dashboard. The full key is shown once.
Sign in
Request builder
Build a chat completion request
Target
Runs on your computer with koldos serve. This request works today against a running KOLDOS.
Sent as the user message. KOLDOS runs one turn for it.
temperature, max_tokens, top_p are accepted and ignored. KOLDOS sets its own budgets for every turn.
curl http://127.0.0.1:8128/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "koldos",
"messages": [
{
"role": "user",
"content": "Explain what the function main does in src/app.py."
}
]
}'This page never sends requests or asks for a key. Run the snippet from your own environment.
Developer portal
Local API at http://127.0.0.1:8128/v1. The hosted API is planned at , which is not deployed.
- 01
API overview
What the API is, how it differs from KOLDOS Local, and the current state of the specification.
Overview - 02
Authentication
An API key in the
AuthenticationAuthorization: Bearerheader. Dashboard sign-in uses Discord through the KOLDOS backend. - 03
API keys
Created and revoked in the dashboard. The full key is shown once and never stored by this site.
Managing keys - 04
Models
Listed per key by
ModelsGET /models. Which models the API serves: Pending. Not published yet - 05
Credits
Each use costs a fixed number of credits: text 1, search 2, vision 3, deep research 8.
Credits - 06
Usage
Daily requests, tokens and credits, plus a per-request log, in the dashboard.
Usage and logs - 07
Rate limits
Per plan, reported in response headers. Values: Pending. Not published yet
Rate limits - 08
Errors
A JSON error object with a stable
Error codescodeand an HTTP status. - 09
Examples
Requests in curl, JavaScript and Python, built from the parameters you choose.
Examples - 10
SDK information
No official SDKs have been published. Any HTTP client works.
SDKs - 11
Changelog
No API changes have been published. The specification is still a draft.
Changelog - 12
Status
Health of the API, authentication, model services and this portal, when a status feed is connected.
Status page
Local endpoints
Available- GET/v1/healthModel, context window and cache type. Open even when a token is set.
- GET/v1/modelsLists the model, served as "koldos".
- POST/v1/chat/completionsA full KOLDOS turn in the common chat completions format, with optional streaming.
- POST/v1/turnsA turn with the KOLDOS-specific result, including route and trace.
- GET/v1/capabilitiesWhat this build implements, including vision availability.
Hosted endpoints
PlannedService status
Status pagePreviewNo status feed is connected, so no health is shown.
- No data. No status feed is connected yet.
API
- No data. No status feed is connected yet.
Authentication
- No data. No status feed is connected yet.
Model services
- No data. No status feed is connected yet.
Developer portal
No live status feed is connected. Account service on this site: Not live.
Plans
PricingFree, Premium, Developer, API. Prices are set; access to the hosted service is granted by hand.
Keys, usage and credits live in the dashboard
Sign in with Discord to manage them once your access is approved.