1stTop pick
Researched todayNex-N2.5-mini
- MMMU
- —
- Output (from)
- Free
Best Multimodal LLMs for Vision (2026)
Last refreshed 2026-09-09. Next refresh: weekly.
Best vision and multimodal LLMs in 2026, ranked by image benchmarks. Covers image QA, document understanding, and video analysis.
Verdict
Use Nex-N2.5-mini for vision today.
Muse Spark 1.3 is the runner-up: — vs — on MMMU.
2ndShortlist
Researched 7d agoMuse Spark 1.3
- MMMU
- —
- Output (from)
- $4.25 / 1M
3rdShortlist
Researched 7d agoMuse Spark 1.3 Contributor
- MMMU
- —
- Output (from)
- $0.200 / 1M
How we rank
Vision/multimodal leaders rank on standard MMMU, use MathVista only as a comparable near-tie signal, then recency. MMMU Pro is tracked separately as harder multimodal evidence, but models without standard MMMU stay benchmark-pending for this leaderboard.
- Eligibility — Models flagged `vision` or `multimodal` in seed data, excluding audio, speech, image-generation, and video-generation specialist models.
- Primary ranking — Standard MMMU score is the primary score. When two scored models are within 1 point and both have MathVista, MathVista breaks the near-tie before release recency.
- Benchmark pending — Recent source-backed vision models without tracked standard MMMU stay visible in a separate benchmark-pending section. MMMU Pro rows remain informational unless the methodology is changed to accept that harder benchmark as a proxy.
- Variant collapse — We keep one row per model family (`familySlug` + parameter tier). When headline scores tie within ±0.5 pt (±10 Elo on Chatbot Arena), we pick the canonical SKU by lowest tracked input price, then GA over preview or limited access, then newest `release`. A folded sibling within the benchmark noise band can show a "Tied within margin" chip on that score cell.
- Pricing — Multimodal pricing often differs by modality — use provider rows for image/video-specific tiers.
| # | Model | Input $/1M | Output $/1M | |
|---|---|---|---|---|
| 1 | Qwen3.6-Plus VisionTools MMMU: 86% | $0.33 | $1.95 | |
| 2 | ByteDance Doubao Seed 2.0 Pro VisionTools MMMU: 85.4% | $0.47 | $2.37 | |
| 3 | Qwen3.5-397B-A17B ReasoningVisionTools MMMU: 85% | $0.39 | $2.34 | |
| 4 | Gemini 3.5 Flash ReasoningVisionTools MMMU: 83.6% | $1.50 | $9.00 | |
| 5 | Claude Sonnet 4.6 ReasoningVisionTools MMMU: 83.6% | $3.00 | $15.00 | |
| 6 | o3 ReasoningVisionTools MMMU: 82.9% | $2.00 | $8.00 | |
| 7 | GPT-5.4 ReasoningVisionTools MMMU: 82.1% | $2.50 | $15.00 | |
| 8 | Qwen3.6 Max Preview PreviewReasoningVisionTools MMMU: 82% | $1.04 | $6.24 | |
| 9 | Gemini 2.5 Pro ReasoningVisionTools MMMU: 81.7% | $1.25 | $10.00 | |
| 10 | Gemini 3 Pro VisionTools MMMU: 81% | $1.25 | $5.00 | |
| 11 | Claude Opus 4.5 ReasoningVisionTools MMMU: 80.7% | $5.00 | $25.00 | |
| 12 | Gemini 2.5 Flash VisionTools MMMU: 79.7% | $0.30 | $2.50 | |
| 13 | Gemini 2.5 Pro Preview 05-06 PreviewVision MMMU: 79.6% | $1.25 | $10.00 | |
| 14 | Claude Sonnet 4.5 ReasoningVisionTools MMMU: 77.8% | $3.00 | $15.00 | |
| 15 | Claude Opus 4.6 ReasoningVisionTools MMMU: 76.5% | $5.00 | $25.00 | |
| 16 | Command A+ ReasoningVisionTools MMMU: 75.1% | — | — | |
| 17 | Claude 3.7 Sonnet ReasoningVisionTools MMMU: 75% | $3.00 | $15.00 | |
| 18 | Llama 4 Maverick 17B Instruct FP8 Vision MMMU: 73.4% | $0.15 | $0.60 | |
| 19 | Llama 4 Scout 17B-16E Instruct Vision MMMU: 69.4% | $0.08 | $0.22 | |
| 20 | GPT-4o VisionTools MMMU: 69.1% | $2.50 | $10.00 |
New models awaiting benchmark coverage
These source-backed rows qualify for this task page, but they are not scored leaderboard picks until the category benchmark data exists.
| Model | Why it is listed | Status | Tracked price |
|---|---|---|---|
| Nex-N2.5-mini Tools | Nex-N2.5-mini reports source-backed vision or multimodal capability; keep it separate from the scored vision ranking until standard MMMU data lands. | Benchmark pending No tracked standard MMMU score yet. | In Free / Out Free |
| Gemini 3.8 Flash ToolsCode execution | Gemini 3.8 Flash reports source-backed vision or multimodal capability; keep it separate from the scored vision ranking until standard MMMU data lands. | Benchmark pending No tracked standard MMMU score yet. | In $0.75 / Out $3.75 |
| Muse Spark 1.3 Tools | Muse Spark 1.3 reports source-backed vision or multimodal capability; keep it separate from the scored vision ranking until standard MMMU data lands. | Benchmark pending No tracked standard MMMU score yet. | In $1.25 / Out $4.25 |
| Muse Spark 1.3 Contributor Tools | Muse Spark 1.3 Contributor reports source-backed vision or multimodal capability; keep it separate from the scored vision ranking until standard MMMU data lands. | Benchmark pending No tracked standard MMMU score yet. | In $0.10 / Out $0.20 |
| Qwen3.8-Max-0902 Tools | Qwen3.8-Max-0902 reports source-backed vision or multimodal capability; keep it separate from the scored vision ranking until standard MMMU data lands. | Benchmark pending No tracked standard MMMU score yet. | In $2.00 / Out $6.00 |
| Claude Opus 4.8 ToolsCode execution | Claude Opus 4.8 reports source-backed vision or multimodal capability; keep it separate from the scored vision ranking until standard MMMU data lands. | Benchmark pending No tracked standard MMMU score yet. | In $5.00 / Out $25.00 |
| GPT-5.5 ToolsCode execution | GPT-5.5 reports source-backed vision or multimodal capability; keep it separate from the scored vision ranking until standard MMMU data lands. | Benchmark pending No tracked standard MMMU score yet. | In $5.00 / Out $30.00 |
| Claude Opus 4.7 ToolsCode execution | Claude Opus 4.7 reports source-backed vision or multimodal capability; keep it separate from the scored vision ranking until standard MMMU data lands. | Benchmark pending No tracked standard MMMU score yet. | In $5.00 / Out $25.00 |
Honorable mentions
Next seats in this ranking. Lines below are from each model's stored description in LLMReference seed data—spot-check the model page before relying on a capability claim.
Gemini 3.8 Flash is Google DeepMind's generally available most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows, released September 2, 2026. It accepts text, image, video, audio, and PDF inputs and returns text, with a 1,048,576-token context window, up to 65,536 output tokens, and Gemini API support for thinking (low/medium/high; minimal is not supported), function calling, tool use, structured outputs, code execution, prompt caching, search grounding, URL context, computer use (preview), and batch, flex, and priority consumption. Official model ID: gemini-3.8-flash. Sibling of live gemini-3.7-flash under a new family gemini-3.8. Compare it for Coding, Agents, Long context, Vision, and JSON / Tool use.
—
MMMU
Qwen3.8-Max-0902 is Alibaba's dated API snapshot of Qwen3.8-Max, listed September 2, 2026 on QwenCloud and Model Studio as model ID qwen3.8-max-0902 (first-party alias qwen3.8-max-2026-09-02). Official copy: an upgraded snapshot of qwen3.8-max with stronger coding (engineering-scale / long-horizon autonomous development), collaborative-agent multi-tool orchestration, and refined native vision (chart reasoning, document parsing, multimodal perception). Retains the 1,000,000-token context window, thinking mode, and full tool ecosystem. Input image/text/video, output text; function calling, structured outputs, context cache. International QwenCloud / Model Studio Singapore list price $2 input / $6 output per 1M tokens; implicit cache $0.25; explicit cache create $2.50; explicit cache read $0.17. No first-party 0902 weight drop. Sibling of seeded qwen3.8-max under family qwen3.8. Compare it for Agents and Coding.
—
MMMU
Anthropic's generally available Mythos-class model for demanding reasoning, long-horizon agents, and ambitious coding, released September 1, 2026. Same underlying model as Claude Mythos 5.1 with cybersecurity and biology safeguards (flagged queries route to Opus 4.8 for cyber and Opus 5 for biology; those reroutes are not billed at Fable prices). Claude API id claude-fable-5-1. List price $10 per 1M input tokens and $50 per 1M output tokens; cache reads $0.25 per 1M (0.025x base input, 75% less than Fable 5's $1.00). 1M-token context, 128K max output, adaptive thinking always on (default effort high), text and image input to text output, vision, tool use. Available to Pro, Max, Team, and Enterprise users and on the Claude Platform natively, Amazon Bedrock, Google Cloud, Microsoft Foundry, and Claude Platform on AWS. Reliable knowledge cutoff June 2026. Sibling of seeded claude-fable-5 under family claude-fable. Compare it for Agents and Coding.
—
MMMU
Compare Top Picks
Side-by-side comparison of the top picks by price, benchmark, and API access.
Browse Other Categories
Best LLMs for Code GenerationBest LLMs for RAGBest LLMs for AI Agents (2026): Benchmarks & PricingBest LLMs for ClassificationBest Open Source LLMsBest LLM for Translation in 2026Best AI Image Models in 2026Best AI Video Models in 2026Best LLMs for Reasoning & MathBest Small Language Models (SLMs)Best LLMs for JSON / Tool useCheapest LLM APIs You Can Call Right NowBest Long Context LLMsBest Mainstream LLM APIs, RankedBest LLMs for EnterpriseBest Free LLMs You Can Use Right NowBest LLMs for WritingBest LLMs for MarketingBest LLMs for Customer Support