NousResearch/hermes-agent

Split GGUF multi-shard pricing fix

Cluster 132779 · hermes cli. Data as of

1 open item. Fix merged in #134116 on 6 Oct 2026. Check it covers your case.

The short answer

0open issues
1open PR
3items grouped
Known?
Yes. 3 related items grouped here, 1 issue already closed.
Fix event
A fix was merged to main in PR #134116. Check it covers your case. No release data is tracked.
Tracker
The maintainers' tracker marks this resolved, deciding PR #134116. That is separate from the fix event.
Still open
0 issues and 1 PR.

A bug caused split GGUF models to be priced by only the first shard, making very large models appear tiny and allowing too many to be admitted. A fix computes the total size as the sum of all shards.

Open issues0

No open issues.

Open PRs1

  • PR#132794fix(local-runtime): price a split GGUF as the sum of its shards

Fixes on record

Merged to main. A stamp counts the issues a fix is linked to.

  • #134116 feat(local-models): run Qwen3.8 Flash Next as UD-IQ4_XS with ggml-org's MTP head

    merged to main

    files: apps/desktop/src/i18n/de.ts, apps/desktop/src/i18n/de_local_models.ts, apps/desktop/src/i18n/en.ts, apps/desktop/src/i18n/en_local_models.ts, apps/desktop/src/i18n/es.ts, apps/desktop/src/i18n/es_local_models.ts, apps/desktop/src/i18n/fr.ts, apps/desktop/src/i18n/fr_local_models.ts, apps/desktop/src/i18n/ja.ts, apps/desktop/src/i18n/ja_local_models.ts, apps/desktop/src/i18n/zh-hant_settings.ts, apps/desktop/src/i18n/zh.ts, apps/desktop/src/i18n/zh_local_models.ts, apps/desktop/src/types/hermes.ts, hermes_cli/local_runtime/catalog.json, hermes_cli/local_runtime/catalog.py, hermes_cli/local_runtime/context_policy.py, hermes_cli/local_runtime/estimator.py, hermes_cli/local_runtime/gguf.py, hermes_cli/local_runtime/growth.py (+13 more)

Closed issues · 1
  • Closed#132779Split GGUF priced by shard 1 only: residency probe sees a 175 GB model as 6.7 GiB, admits 3 giants on a 256 GB box
How this cluster was grouped
  • Link evidence (structural links throughout): every member is connected by structural evidence: shared changed files, or a closing/superseded reference.
  • Pairwise link density 0.67: the share of item pairs that carry a link, counting a structural link in full and a relevance-only link at 0.36. 1.00 means every pair is linked.

Pick up a Lab task

Check whether a fix covers the open issues here, with your own agent. Bring back evidence for maintainers.

Open the Lab

At a glance

Subsystem
hermes cli
Merged PRs
#134116
Closed issues
1
Attempts
0 PRs closed without merging
Fix link checked
6 Oct 2026

For your agent

The same facts, typed.