Topics¶
The documentation grouped by subject, so the explanation, the reference, and the procedure for one subject sit together.
Vocabulary¶
accounting: how a value movement is recorded, projected, and audited.chat: conversations sent in the OpenAI chat-completions shape: what a client sends, what a worker serves, and how the front maps between them.generation: what a generation request asks for, and how those choices shape the result.kudos: the kudos economy: what is earned, charged, and priced.moderation: how prompts are filtered, what the submitter can read back, and what evidence is retained for moderators.operations: procedures an operator runs against a deployment.requests: how generation requests progress, match workers, and expose status.workers: how worker capabilities and availability affect request execution.
By topic¶
accounting¶
- Explanation: Kudos accounting, projection, and concurrency
- Explanation: Kudos event retention
- How-to: Kudos ledger operations
- Reference: Kudos accounting reference
chat¶
- Explanation: Chat requests on the Horde
- How-to: Send chat completions through the OpenAI-compatible front
- How-to: Add chat completions to a web frontend
- How-to: Write a minimal chat worker
- How-to: Serve chat jobs from a text worker
- Reference: Chat wire examples
- Tutorial: Your first chat completion through the Horde
- Tutorial: Serve your first chat job with koboldcpp
generation¶
- Explanation: Request feasibility and queue pressure
- How-to: Add extended image controls to a frontend
- How-to: Add a sampler, scheduler, solver control, or annotator
- How-to: Add an image baseline
- How-to: Use the style endpoints
- How-to: Add style authoring and styled requests to a frontend
- How-to: Update an existing style integration
- How-to: Send text generation requests through the native API
- Reference: Samplers and schedulers reference
- Reference: Image baseline policy reference
- Reference: Image model reference loader reference
- Reference: Style contract reference
kudos¶
- Explanation: Kudos accounting, projection, and concurrency
- Explanation: Kudos event retention
- How-to: Kudos ledger operations
- How-to: Add an image baseline
- Reference: Kudos accounting reference
- Reference: Image baseline policy reference
moderation¶
- Reference: Prompt provenance reference
- Reference: Moderation operations reference
operations¶
- How-to: Kudos ledger operations
- How-to: Run the OpenAI-compatible chat front
- Reference: Moderation operations reference
- Reference: Chat front reference
- Reference: Outbound requests reference
requests¶
- Explanation: Request feasibility and queue pressure
- Explanation: Chat requests on the Horde
- How-to: Use the style endpoints
- How-to: Add style authoring and styled requests to a frontend
- How-to: Update an existing style integration
- How-to: Run the OpenAI-compatible chat front
- How-to: Send chat completions through the OpenAI-compatible front
- How-to: Send text generation requests through the native API
- How-to: Add chat completions to a web frontend
- Reference: Prompt provenance reference
- Reference: Moderation operations reference
- Reference: Style contract reference
- Reference: Chat requests reference
- Reference: Chat wire examples
- Reference: Chat front reference
- Reference: Outbound requests reference
- Tutorial: Your first chat completion through the Horde
workers¶
- Explanation: Request feasibility and queue pressure
- Explanation: Chat requests on the Horde
- How-to: Add an image baseline
- How-to: Write a minimal chat worker
- How-to: Serve chat jobs from a text worker
- Reference: Image baseline policy reference
- Reference: Image model reference loader reference
- Reference: Chat requests reference
- Reference: Chat wire examples
- Tutorial: Serve your first chat job with koboldcpp