Issues

• Context length missing from /v1/models endpoint

  questionsupport
  mostlygeek/llama-swap#1281 · Filter by author coryalder opened on Oct 29, 2026
  11

• server: support v1/realtime

  enhancement
  mostlygeek/llama-swap#1277 · Filter by author mostlygeek opened on Oct 28, 2026
  4

  

• Feature Request: Support hipfire stats

  compatenhancement
  mostlygeek/llama-swap#1269 · Filter by author ltjax opened on Oct 26, 2026
  2

• Feature Request: Gaming Mode - Llama Swap unloads all models, and suspends new incoming requests

  enhancementUIwindows
  mostlygeek/llama-swap#1268 · Filter by author jhemmond opened on Oct 26, 2026
  6

• refactor performance management

  enhancementUI
  mostlygeek/llama-swap#1262 · Filter by author mostlygeek opened on Oct 24, 2026
  4

• Feature Request: Experimental Performance Charts - Add Fan Speed

  unconfirmed bug
  mostlygeek/llama-swap#1260 · Filter by author CircuitGuy opened on Oct 24, 2026

• Display versions on Settings page

  enhancement
  mostlygeek/llama-swap#1254 · Filter by author constgen opened on Oct 21, 2026
  3

  

• ComfyUI is terminated while generating images during a matrix model switch

  mostlygeek/llama-swap#1252 · Filter by author brunocasado opened on Oct 21, 2026
  5

• Feature request: instance redundancy and syncing with load balancing

  mostlygeek/llama-swap#1247 · Filter by author andi-at-1 opened on Oct 20, 2026
  2

• Avoid waking runtime-suspended NVIDIA GPUs for background performance monitoring

  unconfirmed bug
  mostlygeek/llama-swap#1246 · Filter by author henfiber opened on Oct 20, 2026

• Feature request: cache invalidation checker

  enhancement
  mostlygeek/llama-swap#1241 · Filter by author bonswouar opened on Oct 17, 2026
  2

• strategy: warm falls back to targets[0] while that target is stopping

  mostlygeek/llama-swap#1240 · Filter by author dfein38347g opened on Oct 17, 2026
  1

• Playground model picker never opens when a model's alias equals its own id (svelte each_key_duplicate)

  bug
  mostlygeek/llama-swap#1237 · Filter by author sprd-tmu opened on Oct 15, 2026
  1

  

• Feature request: Emulate enough of llama.cpp's API to enable Pi's new "use any chat model as decision model" feature

  mostlygeek/llama-swap#1236 · Filter by author ackalker opened on Oct 15, 2026
  7

• Feature request: globalConcurrencyLimit for /upstream endpoint

  mostlygeek/llama-swap#1235 · Filter by author slu-010101 opened on Oct 15, 2026

• Feature request: cli argument to print out rendered llama-swap config.

  mostlygeek/llama-swap#1234 · Filter by author egidijus opened on Oct 15, 2026
  2

• Startup log prints an invalid URL (http://:8080) when listening on all interfaces

  unconfirmed bug
  mostlygeek/llama-swap#1230 · Filter by author ackalker opened on Oct 14, 2026
  7

• Activity shows unknown Prefill/Decode rates for TensorFold responses

  compat
  mostlygeek/llama-swap#1226 · Filter by author jlave-dev opened on Oct 12, 2026
  5

• bug: prompt and cached token values vary by backend

  bug
  mostlygeek/llama-swap#1218 · Filter by author knguyen298 opened on Oct 10, 2026
  2

• Inconsistent Model IDs Between UI and /api/metrics/activity

  unconfirmed bug
  mostlygeek/llama-swap#1217 · Filter by author kripper opened on Oct 8, 2026
  1

• Half throughput before restart of the service

  supportunconfirmed bug
  mostlygeek/llama-swap#1216 · Filter by author savergiggio opened on Oct 8, 2026
  5

• [bug][ui] Drawer in mobile ui does not automatically dismiss when item is selected

  unconfirmed bug
  mostlygeek/llama-swap#1210 · Filter by author cphlipot opened on Oct 7, 2026

• feature: enable differential reloads

  mostlygeek/llama-swap#1202 · Filter by author mostlygeek opened on Oct 4, 2026
  2

• feature request: configurable proxy paths

  mostlygeek/llama-swap#1193 · Filter by author khimaros opened on Oct 1, 2026
  4

• v259 segfaults on startup on macOS arm64 (SIGSEGV in internal/hw.Detect -> gopsutil getFrequency)

  mostlygeek/llama-swap#1178 · Filter by author guzai2000 opened on Sep 26, 2026
  3