Issues

• [REQUEST] kvarn support

  theroyallab/tabbyAPI#492 · Filter by author rpvelloso opened on Oct 2, 2026

• [BUG] Don't switch port when it is just listened to

  bug
  theroyallab/tabbyAPI#490 · Filter by author erazortt opened on Sep 30, 2026
  1

• [BUG] tool_calls silently null unless tool_format is explicitly configured (no error/warning)

  theroyallab/tabbyAPI#479 · Filter by author cCoolBoox opened on Sep 16, 2026
  3

• [DOCS] Running large MoE models under VRAM pressure: autosplit_reserve sizing, offload requirements, and worker tuning

  theroyallab/tabbyAPI#477 · Filter by author sorasoras opened on Sep 12, 2026

• [REQUEST] exllamav3 backend: expose the missing MoE offload config surface (moe_cpu_split, draft_moe_cpu_layers)

  theroyallab/tabbyAPI#476 · Filter by author sorasoras opened on Sep 12, 2026
  1

• Stream partial tool_calls argument deltas for coding agents (parity with vLLM/llama.cpp)

  theroyallab/tabbyAPI#460 · Filter by author tengotengo opened on Sep 2, 2026

• [BUG] SSE delivery is throttled to ~11 tokens/s regardless of generation speed; the response accumulates and flushes when generation ends

  bug
  theroyallab/tabbyAPI#457 · Filter by author raikensan opened on Aug 24, 2026
  2

• [REQUEST] I made a launcher for TabbyAPI

  theroyallab/tabbyAPI#455 · Filter by author SaSteMi opened on Aug 20, 2026

• [REQUEST] Expose llama.cpp-style top-level timings and cached/draft token stats in OAI responses

  theroyallab/tabbyAPI#454 · Filter by author dfein38347g opened on Aug 19, 2026
  1

• [REQUEST] (bug?) Potential MTP prefill scheduling optimisation: deferred cache population improves long-context throughput ~43%

  theroyallab/tabbyAPI#451 · Filter by author Britlion opened on Aug 17, 2026

• [REQUEST] Clarify sampler override file search mechanism

  theroyallab/tabbyAPI#419 · Filter by author klement opened on May 1, 2026

• [REQUEST] Can't Set Context Length via front-end

  theroyallab/tabbyAPI#400 · Filter by author zapshe opened on Dec 6, 2025
  12

• [BUG] Default docker /dev/shm size too small for tensor parallelism to work

  bug
  theroyallab/tabbyAPI#387 · Filter by author robfuscator opened on Oct 11, 2025
  10

• [REQUEST] Enabling support for vision draft models.

  theroyallab/tabbyAPI#384 · Filter by author MoetasimR opened on Oct 5, 2025

• [BUG] "backend:cudaMallocAsync" cause OOM

  bug
  theroyallab/tabbyAPI#381 · Filter by author yodiaditya opened on Sep 5, 2025

• [BUG] Model load fails often but then loads after 3-4 attempts

  bug
  theroyallab/tabbyAPI#319 · Filter by author ncoder-ai opened on Apr 9, 2025

• [BUG] Model Switching Fails with Concurrent Requests

  bug
  theroyallab/tabbyAPI#304 · Filter by author bold84 opened on Mar 14, 2025

• [BUG] JSON schema Decoding exception, the output result is inconsistent with the schema definition

  bugexternal library
  theroyallab/tabbyAPI#297 · Filter by author ZanePoe opened on Mar 4, 2025
  1

• [BUG] RuntimeError: [_efficient_attention_forward] Unsupported mask type on ROCM, for now

  bug
  theroyallab/tabbyAPI#284 · Filter by author ZhenyaPav opened on Feb 14, 2025
  1