Best LLMs for Multimodal & Grounded — October 2026 Leaderboard
Data refreshed:
Vision, document, and grounded enterprise workflow benchmarks
As of October 2026, the top multimodal & grounded model on the BenchLM leaderboard is Kimi K3.1 Pro with a weighted multimodal & grounded score of 98.
Decision lens: use provisional-ranked mode for broader public evidence and verified-ranked mode for source-only comparisons. A model can move between views as evidence coverage changes.
- Data refreshed
- October 23, 2026
- Provisional-ranked
- 56 of 992 models
- Verified-ranked
- 56 of 992 models
- Weighted evidence
- 4 of 32 benchmarks
32 tracked benchmarks
MMMU-Pro, OCRBench V2, olmOCR, VoxPopuli WER, OfficeQA Pro, MMMU-Pro w/ Python, OmniDocBench 1.5, Liquid Extract JSON Validity, Liquid Extract F1, Liquid Extract VLM Judge, GDPval-AA, Blueprint-Bench 2, MedXpertQA (MM), ZeroBench, Design2Code, Flame-VLM-Code, Vision2Web, ImageMining, MMSearch, MMSearch-Plus, SimpleVQA, Facts-VLM, V*, BabyVision, CharXiv (overall), A-OKVQA, OmniDocBench 1.6, ParseBench, ExtractBench (LlamaIndex), ExtractBench Pro
Scope: Vision, Document-office, GUI/web, Video
Evidence set: MMMU-Pro, OCRBench V2, olmOCR, VoxPopuli WER, OfficeQA Pro, MMMU-Pro w/ Python, OmniDocBench 1.5, Liquid Extract JSON Validity, Liquid Extract F1, Liquid Extract VLM Judge, GDPval-AA, Blueprint-Bench 2, MedXpertQA (MM), ZeroBench, Design2Code, Flame-VLM-Code, Vision2Web, ImageMining, MMSearch, MMSearch-Plus, SimpleVQA, Facts-VLM, V*, BabyVision, CharXiv (overall), A-OKVQA, OmniDocBench 1.6, ParseBench, ExtractBench (LlamaIndex), ExtractBench Pro
Scope: Vision, Document-office, GUI/web, Video
Best Multimodal & Grounded picks
BenchLM summaries for multimodal & grounded plus the practical tradeoffs users check next: open weights, price, speed, latency, and context.
Multimodal & Grounded Leaderboard
Primary score: weighted multimodal & grounded score. Higher values rank first. Use the Show metric control to change the value shown in each row.
Switch between provisional-ranked and verified-ranked modes to compare the broader public dataset with sourced-only rankings.
Filters
| Parameters (B) | ||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 Kimi K3.1 Pro Moonshot AI | Not reported | Pending | 98% | Est.82 | 84.1% | — | — | — | 67.4% | 90% | — | — | — | — | — | — | — | 39.0% | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
2 | Not reported | Open | 92% | 81 | 82.3% | 74.2% | — | — | — | — | 92.1% | — | — | — | — | — | 80.4% | 24.0% | — | — | 69.0% | — | — | — | 75.0% | — | — | 82.0% | — | — | — | — | — | — |
3 Claude Opus 5.5 Anthropic | Not reported | Closed | 91.8% | 82 | — | — | — | — | 67.7% | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 79.85% | 86.06% | 51% |
4 Claude Opus 5 Anthropic | Not reported | Closed | 90.9% | 88 | — | — | — | — | 66.9% | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
5 Kimi K3 Moonshot AI | Not reported | Pending | 90.2% | 78 | 81.6% | — | — | — | 63.3% | 83.4% | — | — | — | — | — | — | — | 23.0% | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 83.17% | — |
6 GPT-5.6 Sol OpenAI | Not reported | Closed | 90.1% | 79 | 83% | — | — | — | — | 84.6% | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 75.35% | 93.77% | — |
7 Gemini 3.8 Flash Google | Not reported | Closed | 88.2% | 77 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 72.08% | 80.71% | — |
8 Claude Opus 4.8 Anthropic | Not reported | Closed | 88.1% | 76 | — | — | — | — | 66.2% | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 63.70% | 87.09% | — |
9 Gemini 3.5 Flash Google | Not reported | Closed | 87.4% | 70 | 83.6% | — | — | — | — | — | — | — | — | — | — | 33.6% | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 74.02% | 79.84% | — |
10 | Not reported | Open | 86.1% | 64 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 64.0% | — | — | — | — | — | — | — | — | — | — | — | 89.88% | — |
11 Gemini 3.7 Flash Google | Not reported | Closed | 85.8% | 77 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 71.27% | — | — |
12 GPT-5.6 Terra OpenAI | Not reported | Closed | 84.9% | 75 | 80.7% | — | — | — | — | 82% | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 68.39% | 92.26% | — |
13 Muse Spark Meta | Not reported | Closed | 83.8% | 62 | 80.4% | — | — | — | — | — | — | — | — | — | — | — | 78.4% | 33.0% | — | — | — | — | — | — | 71.3% | — | — | — | — | — | — | — | — | — |
14 | Not reported | Open | 82.8% | 73 | — | — | — | — | — | — | 91.1% | — | — | — | — | — | — | — | — | — | 62.9% | — | — | — | — | — | — | 65.7% | — | — | — | 70.79% | 89.75% | — |
15 Qwen3.7 Plus Alibaba | Not reported | Closed | 81.5% | 71 | 79% | 70.7% | — | — | — | — | 91.4% | — | — | — | — | — | 71.0% | — | — | — | — | — | — | 41.4% | 81.7% | — | — | — | — | — | — | — | — | — |
16 | Not reported | Open | 81% | 68 | — | — | — | — | 62.4% | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 53.4% | — | — | — | 70.75% | 80.75% | — |
17 Claude Sonnet 5 Anthropic | Not reported | Closed | 80.6% | 75 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 62.13% | — | — |
18 Gemini 3.1 Pro Google | Not reported | Closed | 80.3% | 71 | 83.9% | — | — | — | — | — | — | — | — | — | — | — | 81.3% | 29.0% | — | — | — | — | — | — | 72.4% | — | — | — | — | — | — | 69.14% | — | — |
19 GPT-5.6 Luna OpenAI | Not reported | Closed | 80% | 67 | 78.4% | — | — | — | — | 79.5% | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 68.34% | 91.16% | — |
20 Gemini 3 Pro Google | Not reported | Closed | 79.8% | 68 | 81% | — | — | — | — | — | 88.5% | — | — | — | — | — | — | — | — | — | — | — | — | — | 73.2% | — | 88.0% | — | — | — | — | — | — | — |
21 Grok 4.3 xAI | Not reported | Closed | 79.2% | 68 | 78.1% | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
22 Qwen3.6 Plus Alibaba | Not reported | Closed | 75.9% | 68 | 78.8% | — | — | — | — | — | 91.2%P | — | — | — | — | — | — | — | — | — | — | — | — | — | 67.3%P | — | 96.9% | — | — | — | — | — | — | — |
23 GPT-5.2 OpenAI | Not reported | Closed | 75.5% | 58 | 79.5% | — | — | — | — | — | 85.7% | — | — | — | — | — | — | — | — | — | — | — | — | — | 55.8% | — | 75.9% | — | — | — | — | — | — | — |
24 Claude Opus 4.6 Anthropic | Not reported | Closed | 74.9% | 65 | 77.3% | — | — | — | — | — | — | — | — | — | — | — | 64.8% | — | 77.3%P | 98.8%P | 43.5%P | — | 63.8%P | 25.6%P | 63.2%P | — | 66.5%P | — | — | — | — | 54.07% | — | — |
25 | Not reported | Open | 74.7% | 56 | 76.9% | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 62.40% | — | — |
Top AI models for Multimodal & Grounded — October 2026
As of October 2026, Kimi K3.1 Pro leads the provisional multimodal & grounded leaderboard with a score of 98.0%, followed by Qwen3.8 Max (92.0%) and Claude Opus 5.5 (91.8%). BenchLM is currently showing 56 provisional-ranked models and 56 verified-ranked models in this category.
Scored on 3 of the 4 weighted multimodal & grounded benchmarks.
Scored on 2 of the 4 weighted multimodal & grounded benchmarks.
Scored on 2 of the 4 weighted multimodal & grounded benchmarks.
What changed
Kimi K3.1 Pro ranks #1 at 98.0 on the weighted multimodal & grounded score.
Qwen3.8 Max ranks #2 at 92.0 on the weighted multimodal & grounded score.
Claude Opus 5.5 ranks #3 at 91.8 on the weighted multimodal & grounded score.
Top models by benchmark
Frontier multimodal reasoning benchmark spanning charts, diagrams, tables, and visual academic problems(40% of category score)
Score in Context
What these scores mean
The Multimodal & Grounded leaderboard ranks models by a weighted category score. The overall ranking combines external indices with benchmark evidence rather than fixed category weights. The weighted score is led by MMMU-Pro at 40% and OfficeQA Pro at 25%. A model can know facts in text and still fail when the information is in a chart, screenshot, or spreadsheet — this category measures that gap.
Known limitations
Not all models support image input — text-only models are excluded from this category entirely. OfficeQA Pro and CharXiv coverage is still building, so rankings should be read as a blend of available public evidence rather than a complete visual capability profile. Enterprise-specific document formats (scanned PDFs, handwritten notes) remain under-tested by all benchmarks.
How we weight
The Multimodal & Grounded leaderboard ranks models by a weighted category score. The overall ranking combines external indices with benchmark evidence rather than fixed category weights. It remains important for enterprise copilots and document-heavy workflows where models need to interpret visuals, screenshots, and scanned artifacts.
This category tests whether a model can read the world as it actually appears in products: screenshots, charts, scanned documents, and mixed visual-text artifacts. See the multimodal leaderboard or compare with knowledge benchmarks.
Leaderboards exclude benchmark rows that BenchLM generated from other scores or cloned from reference models. When a weighted benchmark is missing after that filter, the category falls back to the remaining trustworthy public rows instead of filling the gap with synthetic values.
The full scoring rules, freshness handling, and runtime/pricing caveats live on the BenchLM methodology page.
Scroll horizontally to read the full evidence ledger.
| Benchmark | Weight | Status | Description |
|---|---|---|---|
| MMMU-Pro | 40% | Weighted | Frontier multimodal reasoning benchmark spanning charts, diagrams, tables, and visual academic problems |
| OfficeQA Pro | 25% | Weighted | Grounded office and enterprise document benchmark |
| CharXiv | 20% | Weighted | |
| AA-MMMU-Pro | 15% | Weighted | |
| OCRBench V2 | — | Display only | A native OCR benchmark for reading text from images across multilingual scripts, low-quality scans, handwriting, structured layouts, charts, and screenshots. |
| olmOCR | — | Display only | An end-to-end document understanding benchmark over long, layout-rich PDFs with tables, equations, headers, footnotes, and multi-column flows. |
| VoxPopuli WER | — | Display only | A speech-recognition benchmark on the cleaned Artificial Analysis VoxPopuli subset, reported as word error rate where lower is better. |
| MMMU-Pro w/ Python | — | Display only | Tool-augmented MMMU-Pro variant that allows Python assistance during multimodal reasoning |
| OmniDocBench 1.5 | — | Display only | Document understanding benchmark measured by edit distance on complex document extraction tasks |
| Liquid Extract JSON Validity | — | Display only | A display-only Liquid AI extraction metric measuring strict JSON parseability. |
| Liquid Extract F1 | — | Display only | A display-only Liquid AI extraction metric measuring requested-field agreement. |
| Liquid Extract VLM Judge | — | Display only | A display-only Liquid AI extraction metric measuring judged agreement with the source image. |
| GDPval-AA | — | Display only | An evaluation focused on professional domain expertise and task delivery quality in office-style knowledge work. |
| Blueprint-Bench 2 | — | Display only | Agentic spatial reasoning benchmark reported as a normalized score. |
| MedXpertQA (MM) | — | Display only | A clinically grounded multimodal medical multiple-choice benchmark with image inputs. |
| ZeroBench | — | Display only | A multi-step visual reasoning benchmark with pass@5 reporting and optional tool use. |
| Design2Code | — | Display only | Multimodal coding benchmark for turning visual designs into working frontend implementations. |
| Flame-VLM-Code | — | Display only | Vision-language coding benchmark for generating correct code from visual and multimodal inputs. |
| Vision2Web | — | Display only | Benchmark for converting visual references into functional web implementations. |
| ImageMining | — | Display only | Multimodal retrieval and extraction benchmark over image-heavy task settings. |
| MMSearch | — | Display only | Multimodal search benchmark for retrieval and grounded answering across mixed-media inputs. |
| MMSearch-Plus | — | Display only | A harder MMSearch variant for multimodal retrieval and grounded tool-use workflows. |
| SimpleVQA | — | Display only | Visual question answering benchmark focused on straightforward image-grounded understanding. |
| Facts-VLM | — | Display only | Grounded multimodal factuality benchmark for evidence-linked answer correctness. |
| V* | — | Display only | Vision-centric benchmark for high-level multimodal reasoning and perception quality. |
| BabyVision | — | Display only | A multimodal benchmark for fine-grained visual perception and grounded reasoning tasks. |
| CharXiv (overall) | — | Display only | A scientific chart understanding benchmark scored across both the descriptive and reasoning splits rather than the reasoning split alone. |
| A-OKVQA | — | Display only | A visual question answering benchmark whose questions cannot be answered from the image alone and require commonsense or world knowledge. |
| OmniDocBench 1.6 | — | Display only | A document parsing benchmark whose overall score composites edit distance, CDM and TEDS sub-scores across diverse PDF layouts. |
| ParseBench | — | Display only | Scores document parsers and vision-language models on 2,078 human-verified enterprise pages across tables, charts, content faithfulness, semantic formatting and visual grounding, with the listed cost per page. |
| ExtractBench (LlamaIndex) | — | Display only | Scores schema-guided extraction from 370 enterprise documents and 4,869 pages on value F1, record completeness, word and page grounding and cost, split by short, medium and long documents. |
| ExtractBench Pro | — | Display only | A harder companion to ExtractBench: 100 synthetic documents and 2,622 pages across 86 document types that test document conventions, house vocabulary, counting rules, drawn relations and charts, scored by perfect-document rate. |
About Multimodal & Grounded benchmarks
Frontier multimodal reasoning benchmark spanning charts, diagrams, tables, and visual academic problems
Questions
What do multimodal and grounded LLM benchmarks measure?
They measure whether models can reason over images, charts, documents, and office artifacts instead of only plain-text prompts.
Which benchmarks matter most here?
MMMU-Pro is a strong frontier multimodal reasoning benchmark, OfficeQA Pro is useful for grounded document and office-style enterprise workflows, and CharXiv tests chart and figure reasoning.
Why is this category separate from knowledge?
A model can know facts in text and still be weak when the information is embedded in a chart, screenshot, PDF, or spreadsheet. This category measures that gap directly.
Multimodal benchmark updates
Multimodal rankings are heating up. Get the weekly update.
Read a sample issueJoin 5,600+ readers.
One email each week. Unsubscribe anytime.