Tools API

Manage libraries, models, and catalog. These endpoints handle server administration tasks.

Base URL

https://api.getkawai.com/v1

Authentication

When authentication is enabled, include your token in the Authorization header:

Authorization: Bearer API_KEY

Libs

Manage llama.cpp libraries installation and updates.

GET /libs

Get information about installed llama.cpp libraries.

Authentication: Optional when auth is enabled.

Headers

Header	Required	Description
Authorization	No	Bearer token for authentication

Response

Returns version information including arch, os, processor, latest version, and current version.

Content-Type: application/json

Examples

Get library information:

curl -X GET https://api.getkawai.com/v1/libs

---

POST /libs/pull

Download and install the latest llama.cpp libraries. Returns streaming progress updates.

Authentication: Required when auth is enabled. Admin token required.

Headers

Header	Required	Description
Authorization	Yes	Bearer token for admin authentication

Response

Streams download progress as Server-Sent Events.

Content-Type: text/event-stream

Examples

Pull latest libraries:

curl -X POST https://api.getkawai.com/v1/libs/pull

---

Models

Manage models - list, pull, show, and remove models from the server.

GET /models

List all available models on the server.

Authentication: Optional when auth is enabled.

Headers

Header	Required	Description
Authorization	No	Bearer token for authentication

Response

Returns a list of model objects with id, owned_by, model_family, size, and modified fields.

Content-Type: application/json

Examples

List all models:

curl -X GET https://api.getkawai.com/v1/models

---

GET /models/{model}

Show detailed information about a specific model.

Authentication: Optional when auth is enabled.

Headers

Header	Required	Description
Authorization	No	Bearer token for authentication

Response

Returns model details including metadata, capabilities, and configuration.

Content-Type: application/json

Examples

Show model details:

curl -X GET https://api.getkawai.com/v1/models/qwen3-8b-q8_0

---

GET /models/ps

List currently loaded/running models in the cache.

Authentication: Optional when auth is enabled.

Headers

Header	Required	Description
Authorization	No	Bearer token for authentication

Response

Returns a list of running models with id, owned_by, model_family, size, expires_at, and active_streams.

Content-Type: application/json

Examples

List running models:

curl -X GET https://api.getkawai.com/v1/models/ps

---

POST /models/index

Rebuild the model index for fast model access.

Authentication: Required when auth is enabled. Admin token required.

Headers

Header	Required	Description
Authorization	Yes	Bearer token for admin authentication

Response

Returns empty response on success.

Content-Type: application/json

Examples

Rebuild model index:

curl -X POST https://api.getkawai.com/v1/models/index

---

POST /models/pull

Pull/download a model from a URL. Returns streaming progress updates.

Authentication: Required when auth is enabled. Admin token required.

Headers

Header	Required	Description
Authorization	Yes	Bearer token for admin authentication
Content-Type	Yes	Must be application/json

Request Body

Content-Type: application/json

Field	Type	Required	Description
model_url	string	Yes	URL to the model GGUF file
proj_url	string	No	URL to the projection file (for vision/audio models)

Response

Streams download progress as Server-Sent Events.

Content-Type: text/event-stream

Examples

Pull a model from HuggingFace:

curl -X POST https://api.getkawai.com/v1/models/pull \
  -H "Content-Type: application/json" \
  -d '{
    "model_url": "https://huggingface.co/Qwen/Qwen3-8B-GGUF/resolve/main/Qwen3-8B-Q8_0.gguf"
  }'

---

DELETE /models/{model}

Remove a model from the server.

Authentication: Required when auth is enabled. Admin token required.

Headers

Header	Required	Description
Authorization	Yes	Bearer token for admin authentication

Response

Returns empty response on success.

Content-Type: application/json

Examples

Remove a model:

curl -X DELETE https://api.getkawai.com/v1/models/qwen3-8b-q8_0

---

Catalog

Browse and pull models from the curated model catalog.

GET /catalog

List all models available in the catalog.

Authentication: Optional when auth is enabled.

Headers

Header	Required	Description
Authorization	No	Bearer token for authentication

Response

Returns a list of catalog models with id, category, owned_by, model_family, and capabilities.

Content-Type: application/json

Examples

List catalog models:

curl -X GET https://api.getkawai.com/v1/catalog

---

GET /catalog/filter/{filter}

List catalog models filtered by category.

Authentication: Optional when auth is enabled.

Headers

Header	Required	Description
Authorization	No	Bearer token for authentication

Response

Returns a filtered list of catalog models.

Content-Type: application/json

Examples

Filter catalog by category:

curl -X GET https://api.getkawai.com/v1/catalog/filter/embedding

---

GET /catalog/{model}

Show detailed information about a catalog model.

Authentication: Optional when auth is enabled.

Headers

Header	Required	Description
Authorization	No	Bearer token for authentication

Response

Returns full catalog model details including files, capabilities, and metadata.

Content-Type: application/json

Examples

Show catalog model details:

curl -X GET https://api.getkawai.com/v1/catalog/qwen3-8b-q8_0

---

POST /catalog/pull/{model}

Pull a model from the catalog by ID. Returns streaming progress updates.

Authentication: Optional when auth is enabled.

Headers

Header	Required	Description
Authorization	No	Bearer token for authentication

Response

Streams download progress as Server-Sent Events.

Content-Type: text/event-stream

Examples

Pull a catalog model:

curl -X POST https://api.getkawai.com/v1/catalog/pull/qwen3-8b-q8_0

---