1stTop pick
Researched 28d agoClaude Fable 5
- LMArena Vision
- 1309
- Output (from)
- $50.00 / 1M
Best Multimodal LLMs for Vision (2026)
As of 2026-11-14, LLM Reference's top pick for vision is Claude Fable 5 by Anthropic (LMArena Vision 1309, $50.00/1M output tokens), chosen from models that meet the podium eligibility rules. Second pick: Qwen3.8-Max by Alibaba.
Last refreshed 2026-11-14. Next refresh: weekly.
Best vision and multimodal LLMs in 2026, ranked by LMArena Vision overall. Models without an LMArena Vision score are listed below, unranked.
Verdict
Use Claude Fable 5 for vision today.
Qwen3.8-Max is the runner-up: 1309 vs 1300 on LMArena Vision.
2ndShortlist
Researched 28d agoQwen3.8-Max
- LMArena Vision
- 1300
- Output (from)
- $6.00 / 1M
3rdShortlist
Researched 19d agoGPT-6.1 Sol
- LMArena Vision
- 1294
- Output (from)
- $10.00 / 1M
How we rank
Vision models are ranked by LMArena Vision overall only. MMMU is shown for reference and is not used for order. Models without an arena score are listed below, unranked, and are never on the podium.
- Eligibility — Models listed with vision or multimodal input, excluding audio, speech, image-generation, and video-generation specialist models.
- Primary ranking — LMArena Vision overall, highest first. No other score is used to rank, and scores are never combined. MMMU is a reference column only.
- Unranked models — Models without an LMArena Vision score are listed below the ranked models, unranked and never on the podium. They are ordered by MMMU where they have a score, then by newest release. A model here never outranks a model with an arena score. Recent vision models with neither score stay visible in a separate benchmark-pending section.
- 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.
- Podium eligibility — Preview, restricted, retired and retiring models never fill a podium slot. Preview means preview or invite-only access, or a model that is announced or pre-release. Restricted means restricted, partner, private or otherwise unavailable access. Retired means the model is past its retirement date or every provider route is retired. Retiring means the model, or any one of its provider routes, has a retirement date within the next 60 days. These models can still appear in the table below the podium.
- Pricing — Multimodal pricing often differs by modality — use provider rows for image/video-specific tiers.
| # | Model | Input $/1M | Output $/1M | |
|---|---|---|---|---|
| 1 | Claude Fable 5 ReasoningVisionTools LMArena Vision: 1309 | $10.00 | $50.00 | |
| 2 | Qwen3.8-Max ReasoningVisionTools LMArena Vision: 1300 | $2.00 | $6.00 | |
| 3 | Claude Opus 4.7 ReasoningVisionTools LMArena Vision: 1299 | $5.00 | $25.00 | |
| 4 | Claude Opus 4.6 ReasoningVisionTools LMArena Vision: 1299 | $5.00 | $25.00 | |
| 5 | Gemini 3.7 Flash ReasoningVisionToolsVerify pricing LMArena Vision: 1295 | $0.75 | $3.75 | |
| 6 | GPT-6.1 Sol ReasoningVisionTools LMArena Vision: 1294 | $2.00 | $10.00 | |
| 7 | Muse Spark ReasoningVisionTools LMArena Vision: 1294 | — | — | |
| 8 | Claude Opus 5.5 ReasoningVisionTools LMArena Vision: 1293 | $4.00 | $20.00 | |
| 9 | Muse Spark 1.2 PreviewReasoningVisionTools LMArena Vision: 1293 | $1.25 | $4.25 | |
| 10 | Gemini 3.8 Flash ReasoningVisionToolsVerify pricing LMArena Vision: 1291 | $0.75 | $3.75 | |
| 11 | Muse Spark 1.3 ReasoningVisionTools LMArena Vision: 1290 | $1.25 | $4.25 | |
| 12 | Claude Fable 5.1 ReasoningVisionToolsVerify pricing LMArena Vision: 1288 | $10.00 | $50.00 | |
| 13 | Claude Opus 5 ReasoningVisionTools LMArena Vision: 1288 | $5.00 | $25.00 | |
| 14 | GPT-5.6 Sol ReasoningVisionTools LMArena Vision: 1286 | $4.00 | $20.00 | |
| 15 | Claude Opus 4.8 ReasoningVisionTools LMArena Vision: 1286 | $5.00 | $25.00 | |
| 16 | GPT-5.5 ReasoningVisionToolsVerify pricing LMArena Vision: 1286 | $5.00 | $30.00 | |
| 17 | GPT-6 Astra RestrictedReasoningVisionTools LMArena Vision: 1285 | $10.00 | $50.00 | |
| 18 | GPT-5.4 ReasoningVisionTools LMArena Vision: 1284 | $2.50 | $15.00 | |
| 19 | Gemini 3.6 Flash ReasoningVisionTools LMArena Vision: 1282 | $0.75 | $3.75 | |
| 20 | Gemini 3.5 Flash ReasoningVisionToolsVerify pricing LMArena Vision: 1282 | $1.50 | $9.00 |
Unranked vision models
Unranked: no LMArena Vision score yet.
| Model | Input $/1M | Output $/1M | |
|---|---|---|---|
| Qwen3.6-Plus VisionTools MMMU (reference): 86% | $0.33 | $1.95 | |
| Claude Opus 4.5 ReasoningVisionTools MMMU (reference): 80.7% | $5.00 | $25.00 | |
| Claude Sonnet 4.5 ReasoningVisionTools MMMU (reference): 77.8% | $3.00 | $15.00 | |
| Command A+ ReasoningVisionTools MMMU (reference): 75.1% | $0.30 | $1.50 | |
| Claude Sonnet 4 Vision MMMU (reference): 74.4% | — | — | |
| Llama 4 Maverick 17B Instruct FP8 VisionVerify pricing MMMU (reference): 73.4% | $0.15 | $0.60 | |
| GPT-4o VisionTools MMMU (reference): 69.1% | $2.50 | $10.00 | |
| Qwen2.5-VL-72B VisionVerify pricing MMMU (reference): 64.5% | $0.80 | $0.80 | |
| Qwen2-VL-72B-Instruct VisionVerify pricing MMMU (reference): 64.5% | $0.90 | $0.90 | |
| MiniCPM-V 4.6 VisionVerify pricing MMMU (reference): 55.3% | — | — | |
| Phi-4 Reasoning Vision 15B VisionVerify pricing MMMU (reference): 54.3% | — | — | |
| GLM-4V 9B Verify pricing MMMU (reference): 48.3% | $0.05 | $0.25 | |
| Phi 3.5 Vision Instruct VisionVerify pricing MMMU (reference): 43% | — | — | |
| Gemini 4 Hydrogen ReasoningVisionTools LMArena Vision: — | $1.50 | $7.50 | |
| pplx-decider-flash-v1.3 Vision LMArena Vision: — | $0.01 | Free | |
| pplx-decider-v1.3 Vision LMArena Vision: — | $0.02 | Free | |
| Gemini Decision 1 Vision LMArena Vision: — | $0.10 | Free | |
| Qwen4-Max-Preview PreviewReasoningVisionTools LMArena Vision: — | — | — | |
| Nex-N2.6 ReasoningVisionTools LMArena Vision: — | — | — | |
| Claude Fable 5.2 ReasoningVisionTools LMArena Vision: — | $8.00 | $40.00 |
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.
Muse Spark is the first model in Meta's Muse family, developed by Meta Superintelligence Labs (MSL). It is a natively multimodal reasoning model with capabilities including tool-use, visual chain-of-thought reasoning, and multi-agent orchestration. Muse Spark achieves 58% on Humanity's Last Exam and 38% on FrontierScience Research benchmarks, while being competitive with Llama 4 Maverick at over 10x less compute. Available via meta.ai and the Meta AI app; private API preview only — not open-source.
1294
LMArena Vision
Anthropic's Claude Opus 5.5, first model in the Claude 5.5 family, released September 22, 2026. First-party positioning: long-running agentic coding and knowledge work; performs at the level of Claude Fable 5.1 on most work; typical workloads about 40% less cost than Opus 5 at default settings (list input/output 20% below Opus 5; cache reads 60% below). Claude API id claude-opus-5-5. List price $4 per 1M input and $20 per 1M output; cache reads $0.20 per 1M tokens (0.05x base input); 5-minute cache writes $5; 1-hour cache writes $8; Batch API 50% off ($2/$10). Fast mode (Claude API research preview) is $8/$40. 1M-token context, 128K max output (Batch API beta up to 300K with output-300k-2026-03-24), adaptive thinking always on (default effort medium), text and image input to text output, vision, and tool use. Available on Claude for Pro, Max, Team, and Enterprise and on the Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry, and Claude Platform on AWS. Reliable knowledge cutoff June 2026.
1293
LMArena Vision
Muse Spark 1.2 is Meta Superintelligence Labs' proprietary multimodal reasoning model for agentic coding and long-context work, released August 5, 2026. It supports a 1,048,576-token context window and up to 131,072 output tokens through the Meta Model API. No public weights release at launch.
1293
LMArena Vision
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 Agents vs Agent Products (2026)Best 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
Frequently asked questions
What is the best LLM for vision?
LLM Reference's top pick for vision as of 2026-11-14 is Claude Fable 5 by Anthropic, with LMArena Vision 1309. Picks come from models that meet the podium eligibility rules, so the top pick may not lead the full table. Second pick: Qwen3.8-Max by Alibaba (LMArena Vision 1300).
Which ranked LLM for vision has the lowest output price?
Among the ranked models shown on this page, Gemini 3.7 Flash by Google DeepMind has the lowest listed output price: $3.75/1M output tokens, at rank 5.
What is the best open-weight LLM for vision?
Qwen3.8-Max by Alibaba is the highest-ranked open-weight model shown on this page (rank 2, LMArena Vision 1300). License: Qwen3.8-Max License.
How is the vision ranking calculated?
Vision models are ranked by LMArena Vision overall only. MMMU is shown for reference and is not used for order. Models without an arena score are listed below, unranked, and are never on the podium. Preview, restricted, retired and retiring models never fill a podium slot. Preview means preview or invite-only access, or a model that is announced or pre-release. Restricted means restricted, partner, private or otherwise unavailable access. Retired means the model is past its retirement date or every provider route is retired. Retiring means the model, or any one of its provider routes, has a retirement date within the next 60 days. These models can still appear in the table below the podium.
How often is this page updated?
Weekly. Last refreshed 2026-11-14.