Web Clone Bench
Claude Opus 5.5 34.2% leads since Oct 2026 Not in model cards
Agent gets a running reference app (Gmail, Jira, Figma, etc.) and 8 hours to build a clone. Expert-written workflows are replayed as end-to-end tests.
9 apps. Shipping the reference's own bundles scores zero. Environments are available through partnership.
Made by Plato Released Oct 2026 Building whole projects Tasks kept private
Last model added Oct 2026
Scores refreshed 2 Nov 2026
- Level
- Research frontier Researchers, on open problems or tasks that take days
- Scored
- Checked automatically A program checks the result against an answer key, unit tests or the expected final state.
- Answers
- Code A program or patch run against tests.
9
Models scored on its own leaderboard
Claude Opus 5.5 leads with 34.2%
Best score per model from the benchmark's own leaderboard, read on 2 Nov 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.
- 1 Claude Opus 5.5 xhigh 34.2%
- 2 GPT-6 Astra max 17.5%
- 3 Claude Fable 5.1 high 13.3%
- 4 GPT-6.1 Sol max 12.6%
- 5
Gemini 4 Argon medium 8.60% - 6
Muse Wonder high 3.70% - 7 Grok 4.7 low 2.80%
- 8 Muse Spark 1.3 medium 1.50%
- 9 Gemini 3.8 Flash medium 1.00%
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Opus 5.5 | 34.2% | xhigh | Sep 2026 | Official leaderboard |
| 2 | GPT-6 Astra | 17.5% | max | Sep 2026 | Official leaderboard |
| 3 | Claude Fable 5.1 | 13.3% | high | Sep 2026 | Official leaderboard |
| 4 | GPT-6.1 Sol | 12.6% | max | Sep 2026 | Official leaderboard |
| 5 | Gemini 4 Argon | 8.60% | medium | Sep 2026 | Official leaderboard |
| 6 | Muse Wonder | 3.70% | high | Oct 2026 | Official leaderboard |
| 7 | Grok 4.7 | 2.80% | low | Sep 2026 | Official leaderboard |
| 8 | Muse Spark 1.3 | 1.50% | medium | Sep 2026 | Official leaderboard |
| 9 | Gemini 3.8 Flash | 1.00% | medium | Sep 2026 | Official leaderboard |
The best score, 34.2%, costs $652 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.
Google DeepMind Meta OpenAI Anthropic
Cost frontier
One model at each effort level
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| Claude Opus 5.5 | xhigh | 34.2% | $652 | ||
| Claude Opus 5.5 | max | 32.3% | $719 | ||
| Claude Opus 5.5 | high | 30.5% | $547 | ||
| Claude Opus 5.5 | medium | 18.6% | $404 | ||
| GPT-6 Astra | max | 17.5% | $207 | ||
| Claude Fable 5.1 | high | 13.3% | $272 | ||
| Claude Fable 5.1 | xhigh | 13% | $314 | ||
| GPT-6.1 Sol | max | 12.6% | $25 | ||
| GPT-6 Astra | xhigh | 11.8% | $117 | ||
| Claude Fable 5.1 | max | 11.6% | $228 | ||
| GPT-6 Astra | high | 10% | $75 | ||
| GPT-6.1 Sol | high | 9.90% | $14 | ||
| GPT-6 Astra | medium | 9.60% | $46 | ||
| Claude Fable 5.1 | low | 9.20% | $175 | ||
| Claude Fable 5.1 | medium | 9.10% | $207 | ||
| GPT-6.1 Sol | xhigh | 8.60% | $17 | ||
| Gemini 4 Argon | medium | 8.60% | $89 | ||
| GPT-6 Astra | low | 7.80% | $31 | ||
| GPT-6.1 Sol | medium | 6.90% | $5.80 | ||
| Gemini 4 Argon | high | 6.30% | $49 | ||
| GPT-6.1 Sol | low | 4.60% | $2.49 | ||
| Muse Wonder | high | 3.70% | $8.60 | ||
| Gemini 4 Argon | low | 3.30% | $39 | ||
| Muse Wonder | xhigh | 2.90% | $6.81 | ||
| Grok 4.7 | low | 2.80% | $18 | ||
| Grok 4.7 | high | 2.70% | $19 | ||
| Claude Opus 5.5 | low | 2.40% | $17 | ||
| Grok 4.7 | medium | 2.10% | $16 | ||
| Muse Wonder | max | 2.00% | $14 | ||
| Muse Spark 1.3 | medium | 1.50% | $5.46 | ||
| Gemini 3.8 Flash | medium | 1.00% | $2.01 | ||
| Muse Spark 1.3 | low | 0.90% | $1.89 | ||
| Gemini 3.8 Flash | high | 0.50% | $2.72 | ||
| Gemini 3.8 Flash | low | 0.20% | $1.66 | ||
| Muse Spark 1.3 | high | 0.20% | $4.20 |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Anthropic
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
On the benchmark’s own leaderboard
Best so far
No lab model card reports it yet
We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark.
More 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%
- SWE-bench Multilingual In use 300 GitHub-issue tasks across nine programming languages in SWE-bench format. Patch graded by repository tests. Claude Opus 5.5 93.9%
- 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% - 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