# NousResearch/hermes-agent · Compression threshold fixes and per-model overrides

## Open now

- 1 issue, 3 PRs open on GitHub.
- #66177: \[Bug\]: Compression effective threshold (~150K) too high for ≤160K-ctx models — never fires, causing cascading 429 on 100K TPM providers
- PR #24704: feat: support fixed token threshold and per-model compression settings
- PR #38139: feat(agent): support per-model compression overrides
- PR #66249: fix(compression): respect user-configured threshold below 0.50 on small-ctx models

These issues and PRs address bugs where auto-compression fails to trigger on small or limited-context models, and add support for configuring fixed or per-model compression thresholds.

## TL;DR

- 8 member(s): 1 open issue(s), 3 closed issue(s), 3 open PR(s), 0 merged PR(s), 1 PR(s) closed without merging.
- No fix has been merged to main; one or more PRs were attempted (see below).
- **Recommended:** review the prior/competing PR(s) below. A fix was attempted but not merged; your case may extend or revive that work.

## Maintainer resolution verdict

- Internal-tracker verdict over this cluster's issue members: `unresolved`. This is distinct from the fix status above. Weigh it yourself.

## How this cluster was grouped

- Link evidence (`STRUCTURAL_PARTIAL`): some members are connected by structural evidence (shared changed files, or a closing/superseded reference); the rest are joined by an automated relevance judgement over their text.
- Members outside that structural core may belong to a different underlying problem.
- Pairwise link density `0.21`: the share of member 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.

## Open related issues

- #66177: \[Bug\]: Compression effective threshold (~150K) too high for ≤160K-ctx models — never fires, causing cascading 429 on 100K TPM providers

## Closed / duplicate issues

- #18733: Per-model or per-provider compression threshold overrides
- #24695: feat: Support fixed token threshold and per-model compression settings in config.yaml
- #37630: Per-model compression threshold overrides

## Prior & competing PRs

- **#15431: fix(agent): clamp compression threshold below context_length to ensure compression can trigger (#14690)**. closed without merge
  - maintainer direction: category `merged_via`; implication `superseded`; refs `#49998`
- **#24704: feat: support fixed token threshold and per-model compression settings**. open
  - maintainer direction: category `uncategorized`; implication `needs_llm`; refs `#24695`
  - files: `agent/context_compressor.py`, `hermes_cli/config.py`, `hermes_cli/setup.py`, `run_agent.py`, `tests/agent/test_context_compressor.py`
- **#38139: feat(agent): support per-model compression overrides**. open
  - maintainer direction: category `superseded_by_pr`; implication `superseded`; refs `#19110`, `#19110`, `#19110`, `#37630`, `#18733`, `#19110`
  - files: `agent/agent_init.py`, `agent/compression_config.py`, `gateway/run.py`, `tests/agent/test_compression_config.py`, `tests/gateway/test_agent_cache.py`
- **#66249: fix(compression): respect user-configured threshold below 0.50 on small-ctx models**. open
  - maintainer direction: category `superseded_by_pr`; implication `superseded`; refs `61bbc3933091dc418d683aed2928063e88be694e`
  - files: `agent/context_compressor.py`, `agent/transports/codex.py`

## Fix status

- No merged fix on record.
- Subsystem: `agent`