Ramen Bench
Agents draw an interactive bowl of vegan ramen in one HTML file. Static checks verify the constraints and visitors rate the rendered pages.
On November 19, 2026, five runs shared the top visitor rating of 5.0. GPT-6 Astra had four votes and Gemini 4 Neon had two.
- Made by
- Stephen Benjamin
- Released
- November 2026
- Kind
- Web and app front ends
- Level
- University
- Scored
- Ranked by votes
- Answers
- Agent task
- Tasks
- All public
- Measures
- Preference
- Subject
- Computer and information sciences, Arts
- Language
- English
- Interaction
- Single turn
- Refresh
- Live
- Run by
- A host
- In views
- Kept fresh
- Strongest model tested
- GPT-6.2 Astra #1 on the capability index
- Last model added
- November 2026
- Scores refreshed
- November 20th, 2026
5 models share the top score of 5
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.
- 5 Claude Opus 5
- 5
Gemini 3.8 Flash
- 5 GPT-6 Astra
- 5 Claude Sonnet 5.5
- 5
Gemini 4 Neon
- 5 Claude Fable 5.1
- 4 Claude Opus 5.5
- 4 Claude Fable 5.2
- 4 GPT-6.2 Sol
- 4 GPT-5.6 Sol
- 4
Gemini 4 Argon
- 4 GPT-6.2 Astra
- 3 GPT-5.6 Luna
- 3
Gemini 3.7 Flash
- 3
Muse Spark 1.3
- 3 GPT-6.1 Sol
- 3
Gemini 4 Carbon
- 3 GPT-5.6 Terra
- 3 GLM-5.3
- 2 GPT-5.5
- 2 Claude Haiku 5.5
- 2
Muse Wonder
- 2
Gemini 4 Hydrogen
- 1 GPT-6 Sol
- 1 Claude Opus 5 Max 5
- 2
Gemini 3.8 Flash High 5
- 3 GPT-6 Astra XHigh 5
- 4 Claude Sonnet 5.5 Max 5
- 5
Gemini 4 Neon XHigh 5
- 6 Claude Fable 5.1 High 5
- 7 Claude Opus 5.5 Max 4
- 8 Claude Fable 5.2 XHigh 4
- 9 GPT-6.2 Sol Max 4
- 10 GPT-5.6 Sol Max 4
- 11
Gemini 4 Argon High 4
- 12 GPT-6.2 Astra High 4
- 13 GPT-5.6 Luna Max 3
- 14
Gemini 3.7 Flash Low 3
- 15
Muse Spark 1.3 XHigh 3
- 16 GPT-6.1 Sol Max 3
- 17
Gemini 4 Carbon Medium 3
- 18 GPT-5.6 Terra Max 3
- 19 GLM-5.3 High 3
- 20 GPT-5.5 XHigh 2
- 21 Claude Haiku 5.5 Max 2
- 22
Muse Wonder High 2
- 23
Gemini 4 Hydrogen High 2
- 24 GPT-6 Sol Max 1
Source: Official Ramen Bench leaderboard
| # | Model | Score | Setting | Released | Run on | Link | Source |
|---|---|---|---|---|---|---|---|
| 1 | Claude Opus 5 | 5 | Max, Claude Code | Jul 2026 | Sep 2026 | bitbin.de | Official leaderboard |
| 2 | | 5 | High, Antigravity CLI | Sep 2026 | Sep 2026 | bitbin.de | Official leaderboard |
| 3 | GPT-6 Astra | 5 | XHigh, Codex CLI | Sep 2026 | Sep 2026 | bitbin.de | Official leaderboard |
| 4 | Claude Sonnet 5.5 | 5 | Max, Claude Code | Sep 2026 | Sep 2026 | bitbin.de | Official leaderboard |
| 5 | | 5 | XHigh, Antigravity CLI | Nov 2026 | Nov 2026 | bitbin.de | Official leaderboard |
| 6 | Claude Fable 5.1 | 5 | High, Claude Code | Sep 2026 | Sep 2026 | bitbin.de | Official leaderboard |
| 7 | Claude Opus 5.5 | 4 | Max, Claude Code | Sep 2026 | Sep 2026 | bitbin.de | Official leaderboard |
| 8 | Claude Fable 5.2 | 4 | XHigh, Claude Code | Nov 2026 | Nov 2026 | bitbin.de | Official leaderboard |
| 9 | GPT-6.2 Sol | 4 | Max, Codex | Nov 2026 | Nov 2026 | bitbin.de | Official leaderboard |
| 10 | GPT-5.6 Sol | 4 | Max, Codex CLI | Jul 2026 | Sep 2026 | bitbin.de | Official leaderboard |
| 11 | | 4 | High, Antigravity CLI | Sep 2026 | Oct 2026 | bitbin.de | Official leaderboard |
| 12 | GPT-6.2 Astra | 4 | High, Codex | Nov 2026 | Nov 2026 | bitbin.de | Official leaderboard |
| 13 | GPT-5.6 Luna | 3 | Max, Codex CLI | Jul 2026 | Sep 2026 | bitbin.de | Official leaderboard |
| 14 | | 3 | Low, Antigravity CLI | Aug 2026 | Sep 2026 | bitbin.de | Official leaderboard |
| 15 | | 3 | XHigh, Muse Code | Sep 2026 | Sep 2026 | bitbin.de | Official leaderboard |
| 16 | GPT-6.1 Sol | 3 | Max, Codex | Sep 2026 | Oct 2026 | bitbin.de | Official leaderboard |
| 17 | | 3 | Medium, Antigravity CLI | Oct 2026 | Nov 2026 | bitbin.de | Official leaderboard |
| 18 | GPT-5.6 Terra | 3 | Max, Codex CLI | Jul 2026 | Sep 2026 | bitbin.de | Official leaderboard |
| 19 | GLM-5.3 | 3 | High, Claude Code | Aug 2026 | Sep 2026 | bitbin.de | Official leaderboard |
| 20 | GPT-5.5 | 2 | XHigh, Codex CLI | Apr 2026 | Sep 2026 | bitbin.de | Official leaderboard |
| 21 | Claude Haiku 5.5 | 2 | Max, Claude Code | Oct 2026 | Oct 2026 | bitbin.de | Official leaderboard |
| 22 | | 2 | High, Muse Code | Oct 2026 | Oct 2026 | bitbin.de | Official leaderboard |
| 23 | | 2 | High, Antigravity CLI | Nov 2026 | Nov 2026 | bitbin.de | Official leaderboard |
| 24 | GPT-6 Sol | 1 | Max, Codex CLI | Sep 2026 | Sep 2026 | bitbin.de | Official leaderboard |
The best score, 5, costs $1.86 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.
Source: Official Ramen Bench leaderboard
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| Claude Opus 5 | Max | Claude Code | 5 | $16 | 216,578 |
| High | Antigravity CLI | 5 | $1.89 | 218,709 | |
| GPT-6 Astra | XHigh | Codex CLI | 5 | $1.86 | 17,084 |
| Claude Sonnet 5.5 | Max | Claude Code | 5 | $8.42 | 638,374 |
| XHigh | Antigravity CLI | 5 | $3.24 | 127,532 | |
| Claude Fable 5.1 | High | Claude Code | 5 | $5.16 | 63,620 |
| GPT-6 Astra | High | Codex CLI | 5 | $1.40 | 15,210 |
| GPT-6 Astra | Max | Codex CLI | 5 | $2.92 | 29,503 |
| Claude Opus 5.5 | Max | Claude Code | 4 | $27 | 528,776 |
| Claude Fable 5.2 | XHigh | Claude Code | 4 | $20 | 239,495 |
| GPT-6.2 Sol | Max | Codex | 4 | $0.76 | 44,308 |
| GPT-5.6 Sol | Max | Codex CLI | 4 | $0.13 | |
| GPT-5.6 Sol | Ultra | Codex CLI | 4 | $0.06 | |
| Claude Opus 5 | High | Claude Code | 4 | $11 | 138,103 |
| Claude Fable 5.1 | Max | Claude Code | 4 | $19 | 260,113 |
| Claude Fable 5.1 | XHigh | Claude Code | 4 | $8.89 | 135,382 |
| GPT-6 Astra | Ultra | Codex CLI | 4 | $2.45 | 17,310 |
| Claude Opus 5.5 | XHigh | Claude Code | 4 | $8.05 | 209,160 |
| High | Antigravity CLI | 4 | $1.70 | 56,710 | |
| Medium | Antigravity CLI | 4 | 26,878 | ||
| Claude Fable 5.2 | High | Claude Code | 4 | $6.75 | 114,920 |
| Claude Fable 5.2 | Max | Claude Code | 4 | $35 | 430,530 |
| GPT-6.2 Astra | High | Codex | 4 | $0.94 | 14,198 |
| GPT-6.2 Sol | Medium | Codex | 4 | $0.24 | 13,302 |
| GPT-6.2 Astra | XHigh | Codex | 4 | $1.54 | 24,570 |
| GPT-6 Astra | Medium | Codex CLI | 4 | $0.96 | 7,437 |
| GPT-6.2 Astra | Max | Codex | 4 | $3.39 | 53,413 |
| GPT-6.2 Sol | Low | Codex | 4 | $0.13 | 7,017 |
| GPT-6.2 Sol | XHigh | Codex | 4 | $0.48 | 27,711 |
| GPT-5.6 Luna | Max | Codex CLI | 3 | $0.03 | 14,952 |
| GPT-5.6 Sol | High | Codex CLI | 3 | $0.06 | |
| GPT-5.6 Sol | XHigh | Codex CLI | 3 | $0.05 | |
| Claude Opus 5 | Low | Claude Code | 3 | $2.05 | 29,261 |
| Claude Opus 5 | Medium | Claude Code | 3 | $4.21 | 53,447 |
| Claude Opus 5 | XHigh | Claude Code | 3 | $8.82 | 132,846 |
| Low | Antigravity CLI | 3 | $0.19 | 36,764 | |
| Claude Fable 5.1 | Medium | Claude Code | 3 | $2.24 | 25,505 |
| XHigh | Muse Code | 3 | $0.01 | 22,219 | |
| Low | Muse Code | 3 | $0.009 | 10,525 | |
| GPT-6 Astra | Low | Codex CLI | 3 | $0.48 | 4,712 |
| Claude Opus 5.5 | High | Claude Code | 3 | $3.10 | 95,661 |
| Claude Opus 5.5 | Medium | Claude Code | 3 | $1.93 | 57,529 |
| GPT-6.1 Sol | Max | Codex | 3 | $0.64 | 37,392 |
| GPT-6.1 Sol | Ultra | Codex | 3 | 19,863 | |
| GPT-6.1 Sol | XHigh | Codex | 3 | $0.47 | 27,539 |
| Medium | Antigravity CLI | 3 | $1.02 | 58,615 | |
| GPT-6.2 Sol | High | Codex | 3 | $0.34 | 21,570 |
| Claude Fable 5.2 | Medium | Claude Code | 3 | $1.93 | 33,022 |
| GPT-5.6 Terra | Max | Codex CLI | 3 | $0.04 | |
| GLM-5.3 | High | Claude Code | 3 | $3.47 | 67,815 |
| Medium | Antigravity CLI | 3 | $0.27 | 38,597 | |
| Claude Fable 5.2 | Low | Claude Code | 2 | $1.66 | 28,664 |
| GPT-5.5 | XHigh | Codex CLI | 2 | $0.13 | |
| GPT-5.6 Luna | High | Codex CLI | 2 | $0.004 | |
| GPT-5.6 Sol | Low | Codex CLI | 2 | $0.07 | |
| GPT-5.6 Sol | Medium | Codex CLI | 2 | $0.04 | |
| GPT-5.6 Terra | Ultra | Codex CLI | 2 | $0.04 | |
| GPT-5.6 Terra | XHigh | Codex CLI | 2 | $0.05 | |
| GPT-5.6 Terra | Medium | Codex CLI | 2 | $0.02 | |
| High | Antigravity CLI | 2 | $1.24 | 185,139 | |
| Medium | Antigravity CLI | 2 | $0.30 | 38,708 | |
| GLM-5.3 | Max | Claude Code | 2 | $3.59 | 94,525 |
| Claude Fable 5.1 | Low | Claude Code | 2 | $0.88 | 8,973 |
| Low | Antigravity CLI | 2 | $0.07 | 11,580 | |
| High | Muse Code | 2 | $0.01 | 15,130 | |
| Medium | Muse Code | 2 | $0.008 | 11,059 | |
| Claude Opus 5.5 | Low | Claude Code | 2 | $0.39 | 10,435 |
| Claude Sonnet 5.5 | Low | Claude Code | 2 | $0.20 | 11,833 |
| Claude Haiku 5.5 | Max | Claude Code | 2 | $0.89 | 204,504 |
| High | Muse Code | 2 | $0.41 | 7,753 | |
| Low | Antigravity CLI | 2 | $0.78 | 55,124 | |
| High | Antigravity CLI | 2 | $0.44 | 28,476 | |
| GPT-6.2 Astra | Low | Codex | 2 | $0.22 | 3,722 |
| GPT-6.2 Astra | Medium | Codex | 2 | $0.35 | 5,337 |
| Low | Antigravity CLI | 2 | $0.36 | 25,861 | |
| Low | Muse Code | 2 | $0.44 | 8,700 | |
| Max | Muse Code | 2 | $0.73 | 14,485 | |
| Low | Antigravity CLI | 2 | $0.10 | 9,322 | |
| Medium | Antigravity CLI | 1 | $0.19 | 16,805 | |
| GPT-5.5 | High | Codex CLI | 1 | $0.08 | |
| GPT-5.5 | Low | Codex CLI | 1 | $0.13 | |
| GPT-5.5 | Medium | Codex CLI | 1 | $0.06 | |
| GPT-5.6 Luna | Low | Codex CLI | 1 | $0.001 | |
| GPT-5.6 Luna | Medium | Codex CLI | 1 | $0.002 | |
| GPT-5.6 Luna | XHigh | Codex CLI | 1 | $0.003 | |
| GPT-5.6 Terra | High | Codex CLI | 1 | $0.03 | |
| GPT-5.6 Terra | Low | Codex CLI | 1 | $0.02 | |
| GPT-6 Sol | Max | Codex CLI | 1 | $0.34 | 17,409 |
| GPT-6 Sol | Ultra | Codex CLI | 1 | $0.56 | 23,069 |
| Claude Haiku 5.5 | Low | Claude Code | 1 | $0.02 | 21,003 |
| Medium | Muse Code | 1 | $0.48 | 8,458 | |
| XHigh | Muse Code | 1 | $0.56 | 10,792 |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official Ramen Bench leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official Ramen Bench 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
- 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 (24 entries).
- Pass Models scored 24 models we track.
- Pass Newest score The newest is from November 2026, for Gemini 4 Neon.
- Pass Frontier models 9 of the top 10 have a score, led by GPT-6.2 Astra (#1).
- Note Headroom Scored as a rating, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on bitbin.de: 92 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.