Support

Settings

Theme

Explore themes

Mode

Library

Looks borrowed from benchmark sites and products. Themes with a light and a dark version follow your mode.

Bring your own theme

Upload a Priors theme file, or have your agent make one: copy the prompt, tell it the look you want, and upload what it returns. With the Priors skill installed it already knows the format.

Custom themes stay in this browser. They travel inside shared chart links and a page's theme, so others see them as you do.

For the builder, Explore and model pages on this device. Published pages keep the theme their owner picked.

Mode

Benchmarks · agentic

Ramp Accounting Bench

Long-horizon accounting work across realistic financial systems: 137 tasks, 3 attempts each in one shared agent harness (Mercor Archipelago), graded by GPT-5.6 Luna against per-task criteria. All models ran at high reasoning effort. Value is the mean share of criteria met across the three attempts. Cost is per attempt.

18 models · 18 results · updated 2026-10-31T01:02:02.814141Z · labs.ramp.com ↗

Leaderboard

Chart settings
18 of 18 selected
Gemini 4 Argon high

Score vs subscription cost per task

Highlight models

Click a point on the chart to highlight that exact effort; click it again to remove it.

Chart settings
18 of 18 selected

Heavy line: best score at each price. Thin lines join one model's effort levels. Cost uses each model's median subscription plan, or API price with no plan data. 6 more models under All models in settings.

More from this benchmark

10 published measures

Breakdown

darker is better within each column · click a column to sort
Model
Gemini 4 Argon high23.426.316.82.929.4976.439.7
Opus 5.5 high10.919.78.038.7611.731.420.2
Fable 5.1 high12.419.76.577.313.134.318.5
Muse Wonder (Max) high14.619.710.25.119.4958.938.2
GPT-6 Astra high11.719.08.037.310.933.316.8
GPT-6.1 Sol high10.916.85.845.8410.936.117.1
Gemini 3.8 Flash high10.914.65.843.658.7695.394.3
Grok 4.7 high9.4919.05.849.4913.190.835.6
GLM 5.3 high11.721.22.929.4918.356.433.7
Muse Spark 1.3 high9.4916.14.386.5711.765.755.3
Sonnet 5.5 high8.0312.42.924.389.4931.320.5
Gemini 3.7 Flash high9.4915.36.575.848.7657.356.5
Haiku 5.5 high7.313.91.466.5712.449.422.5
Qwen 3.8 Max high12.419.71.467.318.362.039.3
DeepSeek V4.1 Flash high12.416.82.194.3814.670.735.4
Kimi K3 high10.213.91.463.6512.450.227.0
GLM 5.3 Flash high7.313.91.466.5712.443.929.2
GPT-6 Luna high7.310.90.733.6510.251.323.5

All results

18 rows
#ModelEffortRamp Accounting BenchSubscription $ / taskAPI $ / taskTimeOutput tokensHarness
1Gemini 4 Argonhigh61.1%$2.16$3.0317 min86kMercor Archipelago
2Opus 5.5high50.5%$0.020$1.343.4 min20kMercor Archipelago
3Fable 5.1high49.7%$0.54$3.485.7 min23kMercor Archipelago
4Muse Wonder (Max)high48.7%$1.76$2.654.5 min24kMercor Archipelago
5GPT-6 Astrahigh48.7%$0.099$2.315.0 min12kMercor Archipelago
6GPT-6.1 Solhigh46.4%$0.024$0.424.9 min11kMercor Archipelago
7Gemini 3.8 Flashhigh45.7%$0.060$1.478.9 min83kMercor Archipelago
8Grok 4.7high44.1%$0.29$3.1115 min55kMercor Archipelago
9GLM 5.3high43.1%$0.22$2.1637 min140kMercor Archipelago
10Muse Spark 1.3high42.5%$0.78$1.563.8 min31kMercor Archipelago
11Sonnet 5.5high42.5%$0.010$0.672.6 min20kMercor Archipelago
12Gemini 3.7 Flashhigh42.3%$0.11$0.855.9 min44kMercor Archipelago
13Haiku 5.5high42.3%$0.004$0.214.2 min52kMercor Archipelago
14Qwen 3.8 Maxhigh41.2%$0.34$2.8515 min63kMercor Archipelago
15DeepSeek V4.1 Flashhigh40.8%$0.009$0.1212 min62kMercor Archipelago
16Kimi K3high39.2%$0.14$1.5511 min31kMercor Archipelago
17GLM 5.3 Flashhigh35.7%free$0.1817 min112kMercor Archipelago
18GPT-6 Lunahigh34.0%$0.004$0.0515.4 min30kMercor Archipelago