Benchmark Almanac

WebsiteBench

Claude Opus 5.5 26.2% leads since September 2026
Not in model cards

Agents explore reference websites and rebuild their appearance, behavior, and persistent state. Hidden browser tests require functional correctness and screenshot similarity, then report the fraction passed.

The suite contains 70 sites and 46,916 checks. Screenshot checks require normalized RGB SSIM of at least 0.99. Reference sites carry research-use restrictions. Release month remains unconfirmed.

Made by
TIGER-AI-Lab
Kind
Web and app front ends
Level
Expert
Scored
Checked automatically
Answers
Agent task
Tasks
All public
Measures
Capability
Language
English
In and out
Text, Screens, Code → Code, Actions
Interaction
Agentic
Environment
Browser, Terminal
Items from
Real work and use
Refresh
One fixed set
Run by
Anyone
In views
Computer use, Agents
Strongest model tested
Claude Opus 5.5 #5 on the capability index
Last model added
September 2026
Scores refreshed
November 20th, 2026
8
Models scored on its own leaderboard

Claude Opus 5.5 leads with 26.2% among models released by September 2026

Best score per model from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.

The newest model with a score here came out in September 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 Claude Opus 5.5 websitebench-70-20261017 26.2%
  2. 2 GPT-6.1 Sol websitebench-70-20261017 20.2%
  3. 3 Claude Opus 5 websitebench-70-20261017 17.1%
  4. 4 GLM-5.3-Flash websitebench-70-20261017 14.7%
  5. 5 DeepSeek V4.1 Flash websitebench-70-20261017 13.5%
  6. 6 Grok 4.6 websitebench-70-20261017 12.2%
  7. 7 Kimi K3 websitebench-70-20261017 10.6%
  8. 8 Qwen 3.8 Max websitebench-70-20261017 7.30%

Source: Official WebsiteBench leaderboard

#ModelScoreSettingDateSource
1 Claude Opus 5.5 26.2% websitebench-70-20261017 Sep 2026 Official leaderboard
2 GPT-6.1 Sol 20.2% websitebench-70-20261017 Sep 2026 Official leaderboard
3 Claude Opus 5 17.1% websitebench-70-20261017 Jul 2026 Official leaderboard
4 GLM-5.3-Flash 14.7% websitebench-70-20261017 Aug 2026 Official leaderboard
5 DeepSeek V4.1 Flash 13.5% websitebench-70-20261017 Sep 2026 Official leaderboard
6 Grok 4.6 12.2% websitebench-70-20261017 Aug 2026 Official leaderboard
7 Kimi K3 10.6% websitebench-70-20261017 Jul 2026 Official leaderboard
8 Qwen 3.8 Max 7.30% websitebench-70-20261017 Aug 2026 Official leaderboard

Rankings by part

We read 3 parts from the benchmark’s own leaderboard: Request, Browser and System. Each ranks models on one part of the test. Claude Opus 5.5 leads 2 of them, and GPT-6.1 Sol leads the other one. Arrows show how far a model moved from its overall rank among the models we track.

Request

8 models we track have a score here.

  1. 1 Claude Opus 5.5 71.2%
  2. 2 GPT-6.1 Sol 64.3%
  3. 3 Claude Opus 5 58.8%
  4. 4 GLM-5.3-Flash 43.3%
  5. 5 Kimi K3 ↑2 41%
  6. 6 DeepSeek V4.1 Flash ↓1 39.9%
  7. 7 Grok 4.6 ↓1 39.3%
  8. 8 Qwen 3.8 Max 28.7%

The best score, 26.2%, costs $305 a task

Runs on the cost frontier score higher than any run that's cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

Z.ai DeepSeek Anthropic OpenAI Cost frontier One model at each effort level
0% 10% 20% 30% $1.00 $10 $100 $1,000 Cost per task, log scale Claude Opus 5.5, websitebench-70-20261017 GPT-6.1 Sol, websitebench-70-20261017 GLM-5.3-Flash, websitebench-70-20261017 0% 10% 20% 30% $1.00 $10 $100 $1,000 Cost per task, log scale Claude Opus 5.5, websitebench-70-20261017 GPT-6.1 Sol, websitebench-70-20261017

Source: Official WebsiteBench leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Opus 5.5 websitebench-70-20261017 26.2% $305
GPT-6.1 Sol websitebench-70-20261017 20.2% $36
Claude Opus 5 websitebench-70-20261017 17.1% $202
GLM-5.3-Flash websitebench-70-20261017 14.7% $4.88
DeepSeek V4.1 Flash websitebench-70-20261017 13.5% $4.62
Grok 4.6 websitebench-70-20261017 12.2% $75
Kimi K3 websitebench-70-20261017 10.6% $36
Qwen 3.8 Max websitebench-70-20261017 7.30% $9.63

3 models have held number one on the official leaderboard since 2023

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

Moonshot AI Anthropic
Own leaderboard Not tested since Sep 2026

Source: Official WebsiteBench leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On the benchmark’s own leaderboard Best so far
0% 20% 40% 60% 80% 100% Near the ceiling (85%) Claude Opus 5.5 Claude Opus 5 Kimi K3 0% 20% 40% 60% 80% 100% Near the ceiling (85%) Claude Opus 5.5 Claude Opus 5

Source: Official WebsiteBench 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 Coding

All 523 in Coding
  • SWE-bench Verified In use 500 human-validated GitHub issues from 12 Python repos. Agent writes a patch that must pass hidden tests. Claude Opus 4.7 83.5%
  • LiveCodeBench v6 In use Competitive programming problems collected after model cutoffs from LeetCode, AtCoder, Codeforces. Scored pass@1 on hidden tests. DeepSeek-V4-Pro 93.5%
  • SWE-bench Multilingual In use 300 GitHub-issue tasks across nine programming languages in SWE-bench format. Patch graded by repository tests. Claude Fable 5.2 96.6%
  • DeepSWE In use 113 original long-horizon engineering tasks across 91 repos and five languages, written from scratch. Hand-written verifiers accept alternative solutions. GPT-6 Astra 74.1%
  • SciCode In use Scientist-curated research coding problems split into subproblems across natural sciences. Code checked against test cases. Seed2.1 Pro 59.8%
  • Codeforces rating In use Model solves recent Codeforces contest problems. An Elo-style rating is computed from simulated contest performance. DeepSeek V4.1 Flash 3,471

Data audit

All 10 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (8 entries).
  • Pass Models scored 8 models we track.
  • Pass Newest score The newest is from September 2026, for GPT-6.1 Sol.
  • Note Frontier models Only 2 of the top 10 have a score. The strongest tested is Claude Opus 5.5 (#5).
  • Pass Headroom The best is 26.2%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on website-bench.com: 8 rows for models we track 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