Skip to content

Replace unhealthy GLM route with Novita#76

Merged
nedcut merged 1 commit into
mainfrom
agent/glm-novita-route
Jul 18, 2026
Merged

Replace unhealthy GLM route with Novita#76
nedcut merged 1 commit into
mainfrom
agent/glm-novita-route

Conversation

@nedcut

@nedcut nedcut commented Jul 18, 2026

Copy link
Copy Markdown
Owner

What changed

  • replace the unhealthy first-party Z.AI FP8 GLM 5.2 route with the exact Novita FP8 endpoint
  • record the clean replacement smoke and retire the superseded Z.AI manifest entry
  • refresh exact-route pricing, runtime evidence, cost estimates, readiness notes, and leaderboard metadata
  • add a registry regression assertion for the amended route

Why

Repeated full-panel preflights found OpenRouter's pinned z-ai/fp8 endpoint at status -2. The exact dated novita/fp8 endpoint was healthy, supported the frozen lane parameters, and serves the same GLM 5.2 snapshot. No full-panel GLM result existed, so the route could be amended without replacing observed panel data.

The replacement smoke completed 4/4 decisions with zero failures, zero truncations, exact Novita provenance, and $0.009225 spend. The empirical ten-route panel estimate is now $46.7742, with the operator spend ceiling remaining authoritative.

Validation

  • replacement endpoint preflight passed
  • replacement smoke passed: 4/4 decisions, zero failures/truncations
  • complete 10-route no-spend panel preflight passed
  • python3 -m pytest -q — 458 passed
  • Ruff format and lint passed
  • python3 -m gm_bench validate-contract passed
  • web lint and production build passed

Summary by CodeRabbit

  • Updates

    • Updated the published GLM 5.2 listing to use the validated Novita route.
    • Refreshed leaderboard and model registry information to reflect the latest accepted configuration.
  • Documentation

    • Updated publication-readiness notes with the latest validation status, route details, and spend accounting.
    • Added a run record documenting successful replacement-route validation.
  • Cost Estimates

    • Refreshed runtime measurements, pricing observations, and projected benchmark costs.

@coderabbitai

coderabbitai Bot commented Jul 18, 2026

Copy link
Copy Markdown

Review Change Stack

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info
⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: c7d86245-481b-4a9a-8ab9-bf6cbc5b2860

📥 Commits

Reviewing files that changed from the base of the PR and between cec75f8 and 3ed3d3e.

📒 Files selected for processing (9)
  • config/openrouter_pricing_snapshot.json
  • config/sota_v2_models.json
  • config/sota_v2_smoke_manifest.json
  • docs/PUBLISH_READINESS.md
  • docs/run_logs/sota-v2-glm-novita-amendment-2026-07-18.md
  • docs/run_logs/sota-v2-smokes-2026-07-17.md
  • results/analysis/output-budget-cost-estimate.json
  • tests/test_publication.py
  • web/src/data/leaderboard.json

📝 Walkthrough

Walkthrough

The frozen OpenRouter GLM 5.2 route changes from Z.AI FP8 to Novita FP8. Registry, smoke manifest, leaderboard, runtime pricing, cost estimates, publication readiness, run logs, and consistency tests are updated accordingly.

Changes

GLM 5.2 route amendment

Layer / File(s) Summary
Route registry and smoke identity
config/sota_v2_models.json, config/sota_v2_smoke_manifest.json, web/src/data/leaderboard.json, tests/test_publication.py
The GLM 5.2 registry and required smoke identifier now use the Novita FP8 route, with updated manifest artifacts, endpoint metadata, leaderboard identity, and publication assertions.
Runtime pricing and cost estimates
config/openrouter_pricing_snapshot.json, results/analysis/output-budget-cost-estimate.json
Runtime timestamps, observed latency, pricing rates, per-model costs, aggregate budget figures, and observation sources are refreshed for the Novita amendment.
Publication readiness evidence
docs/PUBLISH_READINESS.md, docs/run_logs/*
Publication status, decision history, route-amendment evidence, and smoke campaign spend accounting now describe the accepted Novita route and replacement smoke.

Estimated code review effort: 3 (Moderate) | ~20 minutes

Possibly related PRs

  • nedcut/gm-bench#70: Adds smoke-manifest validation and identity-based publication gating for the same OpenRouter GLM 5.2 route transition.
  • nedcut/gm-bench#75: Updates the revised-lane publication registry validation affected by the Z.AI-to-Novita identity change.
🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name Status Explanation Resolution
Docstring Coverage ⚠️ Warning Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%. Write docstrings for the functions missing them to satisfy the coverage threshold.
✅ Passed checks (4 passed)
Check name Status Explanation
Description Check ✅ Passed Check skipped - CodeRabbit’s high-level summary is enabled.
Title check ✅ Passed The title accurately summarizes the main change: swapping the unhealthy GLM route to the Novita route.
Linked Issues check ✅ Passed Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check ✅ Passed Check skipped because no linked issues were found for this pull request.
✨ Finishing Touches
📝 Generate docstrings
  • Create stacked PR
  • Commit on current branch
🧪 Generate unit tests (beta)
  • Create PR with unit tests
  • Commit unit tests in branch agent/glm-novita-route

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

Comment @coderabbitai help to get the list of available commands.

@nedcut
nedcut marked this pull request as ready for review July 18, 2026 19:58
Copilot AI review requested due to automatic review settings July 18, 2026 19:58

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

@nedcut
nedcut merged commit 0f12f21 into main Jul 18, 2026
13 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants