Vision

🏆 Leaderboard

As of August 20, 2026, Claude Opus 5 is #1 for multimodal image understanding at 89.9%. Ranked by MMMU: college-level questions on diagrams, charts, and images. Best multimodal LLM for reading images, charts, and video. MMMU and CharXiv ranks, not image generators. API price on every row.

Updated August 20, 2026282 models33 providers


Claude Opus 5 Anthropic · Proprietary	89.9%	—	—	—	—	$30.00
Claude Fable 5 Anthropic · Proprietary	89.3%	—	—	—	—	$60.00
Gemini 3.7 Flash Google · Proprietary	89.0%	—	88.7%	—	—	$4.50
GPT-5.6 Sol OpenAI · Proprietary	88.8%	83%	—	—	—	$35.00
Gemini 3.6 Flash Google · Proprietary	88.4%	—	89.4%	—	—	$4.50
GPT-5.5 OpenAI · Proprietary	88.3%	83.2%	—	—	—	$35.00
Gemini 3.5 Flash Google · Proprietary	88.3%	83.6%	84.2%	—	—	$10.50
Gemini 3.1 Pro Google · Proprietary	88.2%	80.5%	—	—	—	$17.50
Kimi K3OSS Moonshot AI · Open Source	88.2%	81.6%	91.3%	—	—	$18.00
Qwen3.8 MaxOSS Qwen · Open Source	88.0%	82.3%	—	—	—	$8.00
Gemini 3 Flash Google · Proprietary	87.6%	81.2%	80.3%	—	86.9%	$3.50
GPT-5.4 OpenAI · Proprietary	87.5%	81.2%	—	—	—	$17.50
GPT-5.2 OpenAI · Proprietary	86.7%	79.5%	82.1%	—	85.9%	$15.75
Muse Spark 1.1 Meta · Proprietary	86.6%	—	88.4%	—	—	$5.50
Claude Opus 4.8 Anthropic · Proprietary	86.6%	—	89.9%	—	—	$30.00
GPT-5.6 Terra OpenAI · Proprietary	86.5%	80.7%	—	—	—	$14.00
Kimi K2.6OSS Moonshot AI · Open Source	86.3%	80.1%	86.7%	—	—	$4.93
Muse Spark 1.2 Meta · Proprietary	86.1%	—	—	—	—	$5.50
Qwen3.6 Plus Qwen · Proprietary	86%	78.8%	81.5%	—	84%	$3.50
Claude Opus 4.7 Anthropic · Proprietary	85.5%	—	91%	—	—	$30.00
GPT-5.1 Thinking OpenAI · Proprietary	85.4%	—	—	—	—	$11.25
GPT-5.1 Instant OpenAI · Proprietary	85.4%	—	—	—	—	$11.25
GPT-5.1 Codex Mini OpenAI · Proprietary	85.4%	—	—	—	—	$2.25
GPT-5.1 OpenAI · Proprietary	85.4%	—	—	—	—	$11.25
GPT-5.6 Luna OpenAI · Proprietary	85.0%	78.4%	—	—	—	$1.40
Qwen3.5-397B-A17BOSS Qwen · Open Source	85%	—	—	—	—	$4.20
GPT-5 OpenAI · Proprietary	84.2%	78.4%	81.1%	—	84.6%	$11.25
Qwen3.5-122B-A10BOSS Qwen · Open Source	83.9%	76.9%	77.2%	—	82%	$3.60
Gemini 3.5 Flash-Lite Google · Proprietary	83.6%	—	76.5%	—	—	$2.80
Grok 4.3 xAI · Proprietary	83.1%	—	—	—	—	$3.75
Claude Sonnet 5 Anthropic · Proprietary	83.0%	—	88.3%	—	—	$12.00
Qwen3.6-27BOSS Qwen · Open Source	82.9%	75.8%	78.4%	—	84.4%	$4.20
o3 OpenAI · Proprietary	82.9%	76.4%	78.6%	86.8%	83.3%	$10.00
Gemini 3.1 Flash-Lite Google · Proprietary	82.5%	76.8%	73.2%	—	84.8%	$1.75
Qwen3.5-27BOSS Qwen · Open Source	82.3%	75%	79.5%	—	82.3%	$2.70
Gemini 2.5 Pro Preview 06-05 Google · Proprietary	82%	—	—	—	83.6%	$11.25
Qwen3.6-35B-A3BOSS Qwen · Open Source · via OpenRouter	81.7%	75.3%	78%	—	83.7%	$1.14
o4-mini OpenAI · Proprietary	81.6%	—	72%	84.3%	—	$5.50
GPT-5 mini OpenAI · Proprietary	81.6%	—	—	—	—	$2.25
Qwen3.5-35B-A3BOSS Qwen · Open Source	81.4%	75.1%	77.5%	—	80.4%	$2.25
MiniMax M3OSS MiniMax · Open Source	81.2%	78.1%	—	—	84.6%	$1.50
Gemini 3 Pro Google · Proprietary	81%	81%	81.4%	—	87.6%	$14.00
Claude Opus 4.5 Anthropic · Proprietary	80.7%	—	—	—	—	$30.00
Claude Haiku 4.5 Anthropic · Proprietary	80.7%	—	—	—	—	$6.00
MiMo-V2.5OSS Xiaomi · Open Source	80%	77.9%	81%	—	—	$0.50
Gemini 2.5 Flash Google · Proprietary	79.7%	—	—	—	—	$2.80
Gemini 2.5 Pro Google · Proprietary	79.6%	—	—	—	—	$11.25
GPT-5.4 mini OpenAI · Proprietary	79.3%	76.6%	—	—	—	$5.25
Grok-3 xAI · Proprietary	78%	—	—	—	—	$18.00
o1 OpenAI · Proprietary	77.6%	—	—	71.8%	—	$75.00

Showing 1–50 of 282 models

Next step

You found the model. Now ship the product.

Auth, billing, and the API layer are already decided. Fork 8 finished AI apps, or have us build the first version with you.

Get the stack — $249 Have us build it with you

The index

All Large Language Models

282 models across 33 providers. Search or jump to a lab — every model page stays linked here.

Cursor

2 models

Composer 2 Composer 2 Fast

xAI

15 models

Grok 4.6 Grok 4.5 Grok 4.3 Grok-4.1 Grok-4.1 Fast Non-Reasoning Grok-4.1 Fast Reasoning Grok-4.1 Thinking Grok 4 Fast Grok Code Fast 1 Grok-4 Fast Non-Reasoning Grok-4 Fast Reasoning Grok-4 Grok-3 Grok-3 Mini Grok-2

DeepSeek

16 models

DeepSeek-V4-Pro-0813 DeepSeek-V4-Flash-0731 DeepSeek-V4-Flash-0423 DeepSeek-V4-Flash-Max DeepSeek-V4-Pro-Max DeepSeek-V3.2 DeepSeek-V3.2 (Non-thinking)DeepSeek-V3.2-Exp DeepSeek-R1-0528 DeepSeek-V3 0324 DeepSeek R1 Distill Llama 70B DeepSeek R1 Distill Qwen 32B DeepSeek-R1 DeepSeek-V3.1 DeepSeek-V3 DeepSeek-V2.5

MiniMax

6 models

MiniMax M3 MiniMax M2.7 MiniMax M2.5 MiniMax M2.1 MiniMax M2 MiniMax M1 80K

AI21 Labs

2 models

Jamba 1.5 Large Jamba 1.5 Mini

Amazon

5 models

Nova 2 Lite Nova 2 Sonic Nova Lite Nova Micro Nova Pro

Baidu

1 models

ERNIE 4.5

ByteDance

2 models

Seed 2.1 Turbo Seed 2.0 Pro

Cohere

3 models

North Mini Code 1.0 Command A+Command R+

IBM

1 models

Granite 3.3 8B Instruct

Inception

1 models

Mercury 2

inclusionAI

1 models

Ling-3.0-flash

LG AI Research

1 models

K-EXAONE-236B-A23B

Meituan

4 models

LongCat-Flash-Lite LongCat-Flash-Thinking-2601 LongCat-Flash-Thinking LongCat-Flash-Chat

Microsoft

5 models

MAI-Code-1.1-Flash Phi 4 Mini Phi-4-multimodal-instruct Phi 4 Phi-3.5-mini-instruct

Moonshot AI

9 models

Kimi K3 Kimi K2.7 Code Kimi K2.7 Code HighSpeed Kimi K2.6 Kimi K2.5 Kimi K2 0905 Kimi K2-Instruct-0905 Kimi K2-Thinking-0905 Kimi K2 Instruct

Nous Research

1 models

Hermes 3 70B

NVIDIA

5 models

Nemotron 3.5 Lightning (30B A3B)Nemotron 3 Ultra (550B A55B)Nemotron 3 Super (120B A12B)Nemotron 3 Nano (30B A3B)Nemotron Nano 9B v2

Poolside

2 models

Laguna S 2.1 Laguna XS 2.1

Sakana AI

1 models

Sakana Namazu

StepFun

1 models

Step-3.5-Flash

Tencent

1 models

Hy3

Thinking Machines

1 models

Inkling-Small

Unisound

1 models

Upstage

1 models

Solar Pro 4

Xiaomi

3 models

MiMo-V2.5-Pro MiMo-V2.5 MiMo-V2-Flash

What is the best multimodal LLM?

As of August 20, 2026, Claude Opus 5 by Anthropic is #1 for multimodal image understanding at 89.9%. Ranked by MMMU: college-level questions on diagrams, charts, and images. This board also tracks MMMU, MMMU-Pro, CharXiv-R, MathVista. Next on the same board: Claude Fable 5 and Gemini 3.7 Flash. Related leaders: Gemini 3.5 Flash on MMMU-Pro at 83.6%; Claude Mythos Preview on CharXiv-R at 93.2%. This vision leaderboard ranks models by MMMU. Scores come from public evals. Prices are the live API rates in the table above.

Sources: MMMU (https://mmmu-benchmark.github.io/); OpenAI API pricing (https://developers.openai.com/api/docs/pricing); Anthropic Claude API pricing (https://platform.claude.com/docs/en/about-claude/pricing); Gemini API pricing (https://ai.google.dev/gemini-api/docs/pricing)

Top 8 for vision. Ranked by MMMU: college-level questions on diagrams, charts, and images. Input and output are dollars per million tokens.
Rank	Model	MMMU	Input /M	Output /M
1	Claude Opus 5	89.9%	$5.00	$25.00
2	Claude Fable 5	89.3%	$10.00	$50.00
3	Gemini 3.7 Flash	89.0%	$0.75	$3.75
4	GPT-5.6 Sol	88.8%	$5.00	$30.00
5	Gemini 3.6 Flash	88.4%	$0.75	$3.75
6	Gemini 3.5 Flash	88.3%	$1.50	$9.00
7	GPT-5.5	88.3%	$5.00	$30.00
8	Gemini 3.1 Pro	88.2%	$2.50	$15.00

Best AI for images vs a vision LLM

Best AI for images is a generation query. This page ranks multimodal LLMs that understand images. If you want pictures out, that is a different tool. If you want a model to read a chart, a screenshot, or a PDF figure, sort MMMU and CharXiv.

What is MMMU?

MMMU is the college-level multimodal exam most labs quote. MMMU-Pro is harder. VideoMMMU is video. A model can be great at photos and weak at plots. CharXiv is the chart check.

Related LLM leaderboards

LLM Leaderboard Coding Math Agentic Reasoning Writing Science Finance Long Context Open Source

Individual benchmark pages

Rank one eval at a time. All LLM benchmarks.

HLE SWE Bench Humanity's Last Exam (No Tools)Humanity's Last Exam (With Tools)Terminal-Bench 2.0 Terminal-Bench Terminal-Bench 2.1 Terminal-Bench 2 Terminal-Bench 3.0 ARC-AGI-1 Verified

Vision FAQ

Who ranks #1 on the Vision leaderboard?

As of August 20, 2026, Claude Opus 5 by Anthropic ranks #1 on MMMU at 89.9%. API pricing is $5.00/M input and $25.00/M output.

What is the best multimodal LLM?

What are the top models on MMMU?

The current MMMU ranking as of August 20, 2026 is 1. Claude Opus 5 at 89.9%; 2. Claude Fable 5 at 89.3%; 3. Gemini 3.7 Flash at 89.0%.

Which vision model is the cheapest?

Llama 3.2 11B Instruct is the cheapest scored model on this vision leaderboard at $0.05/M input and $0.05/M output ($0.10 blended). Claude Opus 5 still leads MMMU at 89.9%.

Should I always pick the #1 MMMU model?

Not automatically. Claude Opus 5 leads MMMU, but a cheaper scored model can be the better production choice if the quality gap is small. Use the table to weigh MMMU against input/output price, context window, and related evals.

How often is the Vision leaderboard updated?

Scores and API prices on this page are refreshed from published evals and provider rates. The snapshot is labeled August 20, 2026. Treat it as a current index, not a one-off blog post.

What is the best vision LLM?

MMMU and MMMU-Pro rank college-level multimodal reasoning. CharXiv ranks scientific charts. VideoMMMU ranks video understanding.

What is MMMU?

MMMU is Massive Multi-discipline Multimodal Understanding. It mixes diagrams, charts, and images across science and engineering.

Do all priced models have vision scores?

No. Text-only models are omitted from these columns. Sort by MMMU to see the models that actually publish a vision eval.

Is the best AI for images the same as a vision LLM?

Usually no. Best AI for images often means image generators (Midjourney, Flux, Imagen). A vision or multimodal LLM reads images, charts, and screenshots. This board is the reader, not the generator.

What is ChatGPT vision?

ChatGPT vision is OpenAI's image-in-chat feature, backed by a GPT multimodal model. Compare that SKU here on MMMU and related evals, then check Gemini and Claude vision rows on the same table.

Do text-only models appear here?

They can exist in the index, but vision columns stay empty. Sort by MMMU to hide models that never published a multimodal score.

You found the model. Now ship the product.

All Large Language Models

OpenAI

Anthropic

Google

Cursor

xAI

DeepSeek

MiniMax

Qwen

AI21 Labs

Amazon

Baidu

ByteDance

Cohere

IBM

Inception

inclusionAI

LG AI Research

Meituan

Meta

Microsoft

Mistral

Moonshot AI

Nous Research

NVIDIA

Poolside

Sakana AI

StepFun

Tencent

Thinking Machines

Unisound

Upstage

Xiaomi

Z AI

What is the best multimodal LLM?

Best AI for images vs a vision LLM

What is MMMU?

Related LLM leaderboards

Individual benchmark pages

Vision FAQ

Who ranks #1 on the Vision leaderboard?

What is the best multimodal LLM?

What are the top models on MMMU?

Which vision model is the cheapest?

Should I always pick the #1 MMMU model?

How often is the Vision leaderboard updated?

What is the best vision LLM?

What is MMMU?

Do all priced models have vision scores?

Is the best AI for images the same as a vision LLM?

What is ChatGPT vision?

Do text-only models appear here?