Vibe-Eval
VIBE-Eval is a hard evaluation suite for measuring progress of multimodal language models, consisting of 269 visual understanding prompts with gold-standard responses authored by experts. The benchmark has dual objectives: vibe checking multimodal chat models for day-to-day tasks and rigorously testing frontier models, with the hard set containing >50% questions that all frontier models answer incorrectly.
Gemini 2.5 Pro Preview 06-05 from Google currently leads the Vibe-Eval leaderboard with a score of 0.672 across 8 evaluated AI models.
What Vibe-Eval measures
Vibe-Eval is a multimodal benchmark that evaluates large language models on multimodal, general, and vision tasks. LLM Stats tracks 8 models on this benchmark, with a maximum possible score of 1. Current average across reported models is 0.6, with the leader reaching 0.7.
Compare leaders on the best AI for multimodal, best AI for general and best AI for vision leaderboards.
Publication
- Paper
- Vibe-Eval: A hard evaluation suite for measuring progress of multimodal language models
- Authors
- Piotr Padlewski, Max Bain, Matthew Henderson, Zhongkai Zhu, and 18 others
- Published
- arXiv
- 2405.02287
Abstract
We introduce Vibe-Eval: a new open benchmark and framework for evaluating multimodal chat models. Vibe-Eval consists of 269 visual understanding prompts, including 100 of hard difficulty, complete with gold-standard responses authored by experts. Vibe-Eval is open-ended and challenging with dual objectives: (i) vibe checking multimodal chat models for day-to-day tasks and (ii) rigorously testing and probing the capabilities of present frontier models. Notably, our hard set contains >50% questions that all frontier models answer incorrectly. We explore the nuances of designing, evaluating, and ranking models on ultra challenging prompts. We also discuss trade-offs between human and automatic evaluation, and show that automatic model evaluation using Reka Core roughly correlates to human judgment. We offer free API access for the purpose of lightweight evaluation and plan to conduct formal human evaluations for public models that perform well on the Vibe-Eval's automatic scores. We release the evaluation code and data, see https://github.com/reka-ai/reka-vibe-eval
Gemini 2.5 Pro Preview 06-05 leads with 67.2%, followed by Gemini 2.5 Pro at 65.6% and Gemini 2.5 Flash at 65.4%.
Progress Over Time
Interactive timeline showing model performance evolution on Vibe-Eval
Vibe-Eval Leaderboard
| Context | Cost | License | ||||
|---|---|---|---|---|---|---|
| 1 | — | — | — | |||
| 2 | Google | — | 1.0M | $1.25 / $10.00 | ||
| 3 | Google | — | 1.0M | $0.30 / $2.50 | ||
| 4 | Google | — | — | — | ||
| 5 | Google | — | — | — | ||
| 6 | Google | — | — | — | ||
| 7 | Google | — | — | — | ||
| 8 | Google | 8B | — | — |
FAQ
Common questions about Vibe-Eval.
More evaluations to explore
Related benchmarks in the same category
A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.
A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.
Massive Multitask Language Understanding benchmark testing knowledge across 57 diverse subjects including STEM, humanities, social sciences, and professional domains
Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions
LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.
Instruction-Following Evaluation (IFEval) benchmark for large language models, focusing on verifiable instructions with 25 types of instructions and around 500 prompts containing one or more verifiable constraints