Benchmark Almanac

Mind2Web-SC

Llama 3.3 70B 93% leads since December 2024
Not in model cards

Models generate guardrails that check a web agent’s proposed actions against user constraints. The benchmark measures whether the guardrail permits or blocks each action correctly.

Made by
Zhen Xiang and colleagues
Released
June 2024
Kind
Web browsing
Level
University
Scored
Checked automatically
Answers
Code
Tasks
All public
Measures
Capability
Language
English
In and out
Text, Code → Code
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
Based on
Mind2Web
Strongest model tested
Llama 3.3 70B #171 on the capability index
Last model added
December 2024
Scores refreshed
November 20th, 2026
5
Entries in its paper, 3 models we track

Llama 3.3 70B leads with 93% among models released by December 2024 Llama 3.3 70B leads with 93% among models released by December 2024

Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.

The newest model with a score here came out in December 2024, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 Llama 3.3 70B 93%
  2. 2 GPT-4 (snapshot unspecified) 90%
  3. 3 Llama 3.1-70B 84.5%
  4. 4 Llama 3-70B 83.5%
  5. 5 LlamaGuard3-8B 51%

Source: The Mind2Web-SC paper

Source: The Mind2Web-SC paper

#EntryScoreSettingDateSource
1 Llama 3.3 70B 93% GuardAgent Dec 2024 Official leaderboard
2 GPT-4 (snapshot unspecified) 90% GuardAgent Official leaderboard
3 Llama 3.1-70B 84.5% GuardAgent Jul 2024 Official leaderboard
4 Llama 3-70B 83.5% GuardAgent Apr 2024 Official leaderboard
5 LlamaGuard3-8B 51% LlamaGuard Official leaderboard
#ModelScoreSettingDateSource
1 Llama 3.3 70B 93% GuardAgent Dec 2024 Official leaderboard
2 Llama 3.1-70B 84.5% GuardAgent Jul 2024 Official leaderboard
3 Llama 3-70B 83.5% GuardAgent Apr 2024 Official leaderboard

3 models have held number one in the benchmark’s paper since its release in June 2024

Lab claims count from the day the card came out, independent runs from the model’s release.

Meta
Benchmark paper Passed 85% in Dec 2024, by Llama 3.3 70B

Source: The Mind2Web-SC paper

Llama 3.3 70B on the official leaderboard passed 85% in December 2024

That was 6 months after its release. By the same age, 15% of the benchmarks from 2024 had passed 85%. All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
In the benchmark’s paper Best so far
0% 20% 40% 60% 80% 100% 2024 2025 Near the ceiling (85%) Llama 3.3 70B Llama 3.1-70B Llama 3-70B 0% 20% 40% 60% 80% 100% 2024 2025 Near the ceiling (85%) Llama 3.3 70B Llama 3.1-70B

Source: Official Mind2Web-SC leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Computer and browser use

All 119 in Computer and browser use
  • BrowseComp In use Hard-to-find facts that need persistent web browsing. Short verifiable answers graded for exact correctness. Claude Fable 5 93.3%
  • ScreenSpot-Pro In use GUI grounding on high-resolution professional-software screenshots. The model must locate the target UI element. GPT-6 Astra 92.7%
  • AndroidWorld In use Agents complete 116 tasks across 20 real Android apps in an emulator. Each task checks the device's system state to judge success. GPT-6 Astra 100%
  • OSWorld 2.1 In use Agents complete 108 long-horizon workflows across 31 self-hosted websites and desktop apps, averaging 318 tool calls. Scored by binary completion and checkpoint partial credit. Claude Opus 5 44.3%
  • WebArena-Verified In use Agents complete 812 audited WebArena tasks on self-hosted websites. Deterministic evaluators check the agent's answer and captured network traces. Muse Spark 1.1 69%
  • OSWorld 2.1 (offline subset) In use Agents complete the 82 OSWorld 2.1 tasks on the official offline list, with no internet access. Scored by checkpoint partial credit and strict pass rate. Claude Fable 5.2 89.3%

Data audit

All 8 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
  • Pass Independent runs Scores from the benchmark’s paper (5 entries).
  • Pass Models scored 3 models we track, plus 2 other entries on its leaderboard.
  • Note Newest score The newest came 23 months ago, in December 2024, for Llama 3.3 70B. That’s expected for a saturated test.
  • Note Frontier models None of the top 10 has a score, which is fine for a retired test. The strongest tested is Llama 3.3 70B (#171).
  • Note Headroom Saturated. The best is 93%, past 85% since December 2024.
Leaderboard import 3 pass
  • Pass Importer We read the board on arxiv.org: 9 rows for models we track and 2 other entries at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check