benchmark on Hacker News

68 titles from 63 authors in the last 30 days, with 66 distinct links and 12 Show HN.

Counts titles, it does not confirm facts. Open the source before concluding anything.

Projects on this term

The score comes from our model.

• Show HN: Self-bench – benchmark coding agents on real-world software · score 7.5

  SelfBench generates private coding-agent benchmarks from a repository's own merged pull requests by reconstructing tasks, writing hidden tests, and validating them through a multi-gate pipeline. It runs agents and models on these tasks to plot accuracy versus cost with a Pareto frontier, filling a real gap between generic public benchmarks and repo-specific performance.

• Show HN: AI SRE Arena, an Open Benchmark for AI SRE Agents on Kubernetes · score 7.3

  Project Arena is a vendor-neutral benchmark for evaluating AI SRE agents on Kubernetes. It deploys a disposable Kubernetes cluster, injects faults across 21 scenarios, captures investigation records from any product, and scores them with a configurable judge. It enables reproducible comparisons between AI-driven incident investigation tools.

• Show HN: Autorouter Benchmark on Various PCBs · score 7.0

  autorouter-bench is a reproducible benchmark for PCB autorouters, using KiCad as an independent referee to score completion, DRC violations, and other metrics on real open-hardware boards. It provides curated board sets (8 core boards and 25 additional public boards) with pinned upstream sources and consistent design rules, enabling fair comparison of routing tools. The project fills a gap in objective evaluation for EDA autorouting and is particularly valuable for hardware engineers and researchers.

• Show HN: Benchmark coding agents on deploying apps · score 7.0

  ai-deploy-bench is a benchmark that measures how well AI coding agents (specifically Claude Code) can build and deploy five types of web apps across five cloud providers. It runs each cell in GitHub Actions, verifies deployments with black-box tests, and publishes results including success rates, human intervention needed, security scores, and costs. This provides a reproducible, end-to-end evaluation of agent capability from code generation to live deployment.

Recent titles

1. How to benchmark mixed-GPU training without fooling yourself
   2026-10-31 · nodeox.com · HN discussion
2. ViBench 2.0: A Benchmark for Long-Horizon Software Building
   2026-10-31 · frontier.replit.com · HN discussion
3. GBench: DeepMind open-sources evaluation suite with 86 benchmark suites
   2026-10-31 · github.com · HN discussion
4. MCP server resource benchmark: RAM, startup time and tool counts for 11 servers
   2026-10-31 · techfuelhq.com · HN discussion
5. Live benchmark of web scraping APIs and residential proxies (2026)
   2026-10-31 · github.com · HN discussion
6. DualChem Benchmark
   2026-10-29 · scale.com · HN discussion
7. We benchmark LLMs at PostHog on practical tasks with real data
   2026-10-29 · posthog.com · HN discussion
8. Logs Ingestion Benchmark: Loki, Elasticsearch, VictoriaLogs and ClickHouse
   2026-10-29 · victoriametrics.com · HN discussion
9. Frontier CS added benchmark for Gemini 4 Argon on Antigravity
   2026-10-27 · twitter.com · HN discussion
10. Almost isn't good enough: Introducing VybBench, our Design Fidelity Benchmark
    2026-10-27 · cocolabs.ai · HN discussion
11. Show HN: I made /localllama Reddit benchmark using AI ofc
    2026-10-27 · aexol.ai · HN discussion
12. Static reasoning can't prove runtime behavior: benchmark of AI PR reviewers
    2026-10-27 · heal.dev · HN discussion
13. Show HN: Autorouter Benchmark on Various PCBs
    2026-10-26 · github.com · HN discussion
14. Typed Tibia-rules simulator targeting WGSL for the OT Agent Arena benchmark
    2026-10-24 · github.com · HN discussion
15. AI Coding Agent Security Benchmark
    2026-10-23 · endorlabs.com · HN discussion
16. New EDA Benchmark (sneak peek) results are live
    2026-10-23 · github.com · HN discussion
17. FrontierCS 2: A Benchmark for Computer Science R&D
    2026-10-23 · frontier-cs.org · HN discussion
18. State of IaaS and Compute 2026: market benchmark and cost analysis
    2026-10-22 · cloudcompare.dev · HN discussion
19. AA-PersonalityBench: AI Personality and Communication Style Benchmark
    2026-10-22 · artificialanalysis.ai · HN discussion
20. How I made a benchmark mined from my own repo that agreed with FrontierCode
    2026-10-21 · stet.sh · HN discussion
21. Elektric – Benchmark your AI requests across 150 models for cost/quality
    2026-10-21 · elektric.ai · HN discussion
22. State of PaaS and Web Hosting 2026: market benchmark and cost analysis
    2026-10-21 · cloudcompare.dev · HN discussion
23. The Honest Raw LLM Code Review Benchmark: 21 Models, 7,980 Reviews
    2026-10-20 · ironbee.ai · HN discussion
24. Open-source benchmark for long-horizon web-app coding
    2026-10-20 · github.com · HN discussion
25. Show HN: AML CYCLE 2, a benchmark for long-term agent memory
    2026-10-20 · x.com · HN discussion
26. PE Bench: Benchmark for private equity analyst work
    2026-10-20 · getcompound.ai · HN discussion
27. State of Managed Databases 2026: market benchmark and cost analysis
    2026-10-19 · cloudcompare.dev · HN discussion
28. Calling all GPUs: help us benchmark SonicBoom v0.0.3
    2026-10-19 · github.com · HN discussion
29. Show HN: Independent LLM benchmark index tracking 22 frontier models
    2026-10-18 · vibecoderjournal.com · HN discussion
30. Needle: Search Engine Benchmark for Agents
    2026-10-18 · keenableai.github.io · HN discussion

All trends