reasoning benchmark
Vending-Bench 2 Leaderboard
Vending-Bench 2 tests longer horizon planning capabilities by evaluating how well AI models can manage a simulated vending machine business over extended periods. The benchmark measures a model's ability to maintain consistent tool usage and decision-making for a full simulated year of operation, driving higher returns without drifting off task.
Updated Aug 17, 2026
Models4
Model coverage4
MetricFinal balance
EvidenceB
Vending-Bench 2 Ranking
Higher final balance ranks better on this benchmark.
Sort by
Rank | Model | Final balance | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAN | Final balance8,017.59 usd | Percentile100.0% | Participants4 | EvidenceC | Evaluated |
| Rank02 | ModelZA | Final balance5,634.41 usd | Percentile66.7% | Participants4 | EvidenceC | Evaluated |
| Rank03 | ModelGO | Final balance5,478.16 usd | Percentile33.3% | Participants4 | EvidenceC | Evaluated |
| Rank04 | ModelGO | Final balance3,635 usd | Percentile0.0% | Participants4 | EvidenceC | Evaluated |
Vending-Bench 2 Highlights
The leading models and scores on this benchmark.
Vending-Bench 2 Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for Vending-Bench 2
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
Ranking basisThis vending-bench 2 AI model leaderboard uses descending final balance in the benchmark's original unit. The leaderboard ranking keeps matched price and speed data separate from benchmark evidence.
- 01ANClaude Opus 4.6Anthropic
- Final balance
- 8,017.59 usd
- Price
- $5.0 input / $25 output per 1M tokens
- Speed
- Up to 17 tok/s via Anthropic
Strengths
- Ranks #1 of 4 compared models
- 100th percentile on this benchmark
- C evidence result
Considerations
- This result measures Vending-Bench 2, not total model capability
- 02ZAGLM-5.1Zhipu AI
- Final balance
- 5,634.41 usd
- Price
- $1.4 input / $4.4 output per 1M tokens
Strengths
- Ranks #2 of 4 compared models
- 67th percentile on this benchmark
- C evidence result
Considerations
- This result measures Vending-Bench 2, not total model capability
- 03GOGemini 3 ProGoogle
- Final balance
- 5,478.16 usd
- Speed
- Up to 90 tok/s via Google
Strengths
- Ranks #3 of 4 compared models
- 33th percentile on this benchmark
- C evidence result
Considerations
- This result measures Vending-Bench 2, not total model capability
- 04GOGemini 3 FlashGoogle
- Final balance
- 3,635 usd
- Price
- $0.50 input / $3.0 output per 1M tokens
- Speed
- Up to 65 tok/s via Google
Strengths
- Ranks #4 of 4 compared models
- 0th percentile on this benchmark
- C evidence result
Considerations
- This result measures Vending-Bench 2, not total model capability
Selection summary
Best AI Models for Vending-Bench 2
Claude Opus 4.6 currently leads Vending-Bench 2 with 8,017.59 usd. It is the top model on this specific benchmark, while the best LLM for the broader task should also be checked against other benchmarks, price and runtime.
Use this leaderboard with the supporting benchmark results and coverage details above. A leaderboard position summarizes the selected ranking signal; it does not replace workload-specific testing.
What is Vending-Bench 2?
What Vending-Bench 2 measures and how its scores work.
Vending-Bench 2 tests longer horizon planning capabilities by evaluating how well AI models can manage a simulated vending machine business over extended periods. The benchmark measures a model's ability to maintain consistent tool usage and decision-making for a full simulated year of operation, driving higher returns without drifting off task.
Scores are shown in usd. This benchmark is not independently verified and has an evidence level of B.
- Family
- Vending-Bench 2
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- Yes
- Evaluation key
- vending-bench-2|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about Vending-Bench 2.
Which model scores highest on Vending-Bench 2?
Claude Opus 4.6 is currently ranked first with 8,017.59 usd.
What does Vending-Bench 2 measure?
Vending-Bench 2 tests longer horizon planning capabilities by evaluating how well AI models can manage a simulated vending machine business over extended periods. The benchmark measures a model's ability to maintain consistent tool usage and decision-making for a full simulated year of operation, driving higher returns without drifting off task.
Is a higher final balance better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
4 model results are currently shown.
Does this benchmark affect the overall score?
Yes. This benchmark can contribute to the current LLMBoard capability score.