GLM 5.2 Ties with Opus in Enterprise Code Eval
rohanpaul_ai · x · 2026-07-11
Databricks conducted evaluations using real internal PRs, tests, and million-line codebases rather than relying solely on public benchmarks.
Results show:
- GLM 5.2 entered Databricks' top capability tier, statistically tying in quality with Claude Opus 4.8.
- The quality-cost Pareto frontier for coding tasks features OpenAI, Anthropic, and open-source models, indicating that achieving cutting-edge performance currently requires a multi-model combination.
- For the same model, simply changing the harness/orchestration method can yield about a 2x cost saving.
- Databricks believes open-source models can now handle the most difficult tasks, utilizing Omnigent upfront for multi-harness and multi-model routing.
More from coding & agent
- General agents burn ~140M tokens/month on LinkedIn outreach; scripts do it 60x cheaper — ChrisHarpon2 · 2026-10-26
- NLI alone rejects 5/6 new agent memories; a 0.45 similarity gate fixes it all — Alternative_Pin9598 · 2026-10-26
- LangChain ships Managed Deep Agents 0.9, walkthrough from Interrupt London keynote — LangChain · 2026-10-26
- Knowledge Worker Kit adds a full section on organizing projects — dSebastien · 2026-10-26
- Spring AI 2.1.0-RC1 adds tool-call confirmation, pausing agent loops before real-data access — therealdanvega · 2026-10-26
- METR red-teams OpenAI's agent monitoring, finds bypasses and gaps — thione · 2026-10-26