Repository Issue Activity (beta)

#1077 [Feature Request]: improve Kubernetes as a model runtime

#1078 Adding a Pacstall build script

#916 (feat) Add an endpoint to query llama-server prompt cache status and map cached prompts to originating requests

#1086 [Feature] Global concurrency semaphore

#1090 vllm-wrapper: hardcoded 300s ResponseHeaderTimeout 502s long generations, with no way to raise it

#1066 Add 120a arch for Unified-cuda for blackwell users

#1085 The help chatbot is unable to enumerate all models due to truncation

#967 [Feature Request] Request Body Pre-processing Hook

#1106 # Capture subsystem retains ~256 MB per pooled zstd encoder (GOMAXPROCS × window), ballooning RSS to multi-GB under concurrent /v1/embeddings load

#969 Show backend-reported progress for in-flight requests

#971 Feature request: per-API-key permissions (scope keys to models/endpoints)

#972 Feature request: attribute activity/usage to the API key that made the request

#1105 internal/router: add capabilities compatiblity package

#1021 hybrid setup — local llama-swap models + remote gateway for frontier models

#1027 [Feature] Add support for tabbyAPI

#1010 [Feature request]: log conversations to jsonl

#1009 [Feature Request] Load API keys from a file (or files)

#1004 [Feature] Memory-constrained scheduling with per-model RAM declarations

#995 [Feature Request] Embeddings in web ui

#970 Add SPL (Swap Policy Language) DSL to llama-swap

#591 Feature Request: When ttl expires, preload model

#590 Feature Request: Add Web UI Section for Configuration / Model Management

#570 [Feature request] Add About / Docs page in the UI (with a link to github)

#1103 [Feature Request] Allow UI authentication to be configurable

#1094 [Feature Request] Show state/progress/speed in Activity - In flight requests.