BlogLLM Benchmark Scores as Published, Side by Side
WebArena: Results as Published
Every result the catalog holds for WebArena: 6 models from 3 developers, read from 2 pages. Each number is as printed by the company named beside it; nothing is ranked or rescored. Models are listed by developer, newest first.
Area Computer UseMaintainer Shuyan Zhou et al. (original WebArena paper); ServiceNow (WebArena-Verified release)Introduced 2023Data version 2026103001, as of 2026-10-30
WebArena tests whether a language-guided agent can complete tasks on realistic, self-hosted websites. The environment provides functional sites in four domains: e-commerce, social forums, collaborative software development and content management, plus tools such as a map and external knowledge sources. WebArena-Verified is the corrected release of the task set: the full set has 812 verified tasks and the Hard subset has 258 tasks. Its evaluators are deterministic and run offline on network trace replay, with LLM-as-a-judge evaluation removed.
WebArena (vision)
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| Gemma 4 26B A4BGoogle, released 2026-04-02 | 38.2% | NVIDIA, NVIDIA Nemotron 3.5 Bolt 74B-A9B model cardchecked 2026-10-19 | - |
| Qwen3.6-35B-A3BAlibaba, released 2026-04-16 | 36.6% | NVIDIA, NVIDIA Nemotron 3.5 Bolt 74B-A9B model cardchecked 2026-10-19 | - |
| Nemotron 3.5 Bolt 74B-A9BNVIDIA, released 2026-10-15 | 50.86% | NVIDIA, NVIDIA Nemotron 3.5 Bolt 74B-A9B model cardchecked 2026-10-19 | Model as printed: Nemotron 3.5 Bolt 74B-A9B-24GB (IQ2) |
| 50.25% | NVIDIA, NVIDIA Nemotron 3.5 Bolt 74B-A9B model cardchecked 2026-10-19 | Model as printed: Nemotron 3.5 Bolt 74B-A9B (BF16) |
WebArena-Verified
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| Qwen3.7-PlusAlibaba, released 2026-06-01 | 55.3% | Alibaba, Qwen3.8-27B model cardchecked 2026-10-15 | - |
| Qwen3.6-27BAlibaba, released 2026-04-22 | 48.8% | Alibaba, Qwen3.8-27B model cardchecked 2026-10-15 | - |
| Qwen3.8-27BAlibaba | 64.8% | Alibaba, Qwen3.8-27B model cardchecked 2026-10-15 | - |
Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.