LLM Benchmarks — Scores, Methodology & Live Rankings

What each benchmark measures, how models score, and the current rankings

Reset

Most tested benchmarks

Ranked by how many models have tracked scores — open a benchmark for its full live ranking.

BLINK 5 models tested · latest result 27.02.2025 71.1% DocVQA 5 models tested · latest result 27.02.2025 96.7% InfoVQA 5 models tested · latest result 27.02.2025 86.2% CountBenchQA 4 models tested · latest result 27.02.2025 93% TextVQA 4 models tested · latest result 27.02.2025 87.9% VQAv2 4 models tested · latest result 27.02.2025 92.7% ChartQA 3 models tested · latest result 27.02.2025 76.3% COCOcap 3 models tested · latest result 27.02.2025 116 idx OKVQA 3 models tested · latest result 27.02.2025 60.2% ReMI 3 models tested · latest result 27.02.2025 44.8%

COCOcap

Multimodal

COCO Captioning - image captioning on MS-COCO dataset.

Active View →

DocVQA

Multimodal

Document Visual Question Answering - understanding text in document images. (val split)

Active View →

InfoVQA

Multimodal

Infographic VQA - answering questions about infographics. (val split)

Active View →

TextVQA

Multimodal

TextVQA - visual question answering requiring reading text in images. (val split)

Active View →

ReMI

Multimodal

ReMI - referring to images in multimodal context.

Active View →

ChartQA

Multimodal

ChartQA - answering questions about charts and plots.

Active View →

VQAv2

Multimodal

Visual Question Answering v2 - general image QA.

Active View →

BLINK

Multimodal

BLINK - multimodal benchmark for visual perception tasks.

Active View →

OKVQA

Multimodal

OK-VQA - outside knowledge visual question answering.

Active View →

TallyQA

Multimodal

TallyQA - visual counting question answering.

Active View →

SpatialSense VQA

Multimodal

SpatialSense VQA - spatial reasoning in visual question answering.

Active View →

CountBenchQA

Multimodal

CountBenchQA - visual counting question answering benchmark.

Active View →

Daily-Omni

Multimodal

Daily-Omni as evaluated on the LFM3-Omni-2B card: Omni table, 'Video + audio understanding' group. Multiple-choice accuracy on video with its soundtrack. No ...

Active View →

WorldSense

Multimodal

WorldSense as evaluated on the LFM3-Omni-2B card: Omni table, 'Video + audio understanding' group. Multiple-choice accuracy on video with its soundtrack. No ...

Active View →

AV-SpeakerBench

Multimodal

AV-SpeakerBench as evaluated on the LFM3-Omni-2B card: Omni table, 'Video + audio understanding' group. Multiple-choice accuracy on video with its soundtrack...

Active View →

OmniBench (overall)

Multimodal

OmniBench (overall) as evaluated on the LFM3-Omni-2B card: Omni table, 'Image + audio understanding' group. Image plus audio clip per question, scored with t...

Active View →

OmniBench (speech)

Multimodal

OmniBench (speech) as evaluated on the LFM3-Omni-2B card: Omni table, 'Image + audio understanding' group. OmniBench subset whose audio is speech (same offic...

Active View →

English ASR WER

Multimodal

English ASR WER as evaluated on the LFM3-Omni-2B card: average word error rate (WER) in percent on the ESB / Open ASR benchmark suite; lower is better.

Active View →

English TTS UTMOSv2

Multimodal

English TTS UTMOSv2 as evaluated on the LFM3-Omni-2B card: UTMOSv2 speech quality score on a 1-5 scale, higher is better.

Active View →

VoiceBench

Multimodal

VoiceBench as evaluated on the LFM3-Omni-2B card: audio input, text output, 0-100 scale, higher is better.

Active View →

Tool calling (internal eval)

Multimodal

Tool calling (internal eval) as evaluated on the LFM3-Omni-2B card: Liquid AI internal voice tool-calling evaluation, audio input, percentage, higher is better.

Active View →

MMAR

Multimodal

MMAR as evaluated on the LFM3-Omni-2B card: MMAR audio understanding and reasoning, percentage, higher is better.

Active View →

ImajevBench

Multimodal

ImajevBench v2.0-lite (mohit67890/imajev-bench on Hugging Face), image-based decision/judgment benchmark. Values reported as percentage. Source: https://hugg...

Active View →

MMRB2

Multimodal

MMRB2, multimodal reasoning benchmark v2. Values reported as percentage. Source referenced on the LiquidAI/d1-2B model card.

Active View →

NaturalBench

Multimodal

NaturalBench, benchmark for evaluating vision-language models on natural yes/no image questions. Values reported as percentage. Source referenced on the Liqu...

Active View →