Use cases › Seeing
Understand images and video
Models that take images or video as input and answer questions, describe scenes or ground objects in them.
Benchmarks that represent this job 18
A model's row on one of these third-party boards is what earns the measured label; the same benchmark name in the model's own card earns author-reported.
MMMUMMMU-ProMathVistaVideo-MMEMMBenchMMStarRealWorldQAMMVetChartQAAI2DHallusionBenchMathVisionMLVULVBenchBLINKVQAv2TextVQARefCOCO
Models 202
Narrow by what you can run and pay for. Rows are ordered measured → author-reported → structural, then by watch-list score; click a column to re-sort. Every chip links to its source.
| Model | Evidence | Params | Context | Price from | Runs on | License | Providers | Measured latency |
|---|---|---|---|---|---|---|---|---|
| gemma-4-26B-A4B-it google | author-reported Author reports MMMU Pro 73.8%Author reports MATH-Vision 82.4% +1 | 25.8B (4B active) | 256K | $0.00 | Single high-end GPU | Apache 2.0 | 6 | 0.9 s · 51 tok/s |
| gemma-4-31B-it google | author-reported Author reports MMMU Pro 76.9%Author reports MATH-Vision 85.6% +1 | 31.3B | 256K | $0.00 | Single high-end GPU | Apache 2.0 | 5 | — |
| DeepSeek-V4.1-Flash deepseek-ai | author-reported Author reports MMMU-Pro (EM) 56.5Author reports RefCOCO-avg ([email protected]) 86.0 +1 | 763B | 1M | $0.02 | Cluster | MIT | 8 | — |
| MiniMax-M3 MiniMaxAI | author-reported Author reports mmmu_pro_standard_10_options 78.1Author reports video-mme-v2 85.4 +1 | 427B | 1M | $0.22 | Cluster | MiniMax Model License (mi… gated-threshold | 7 | — |
| Kimi-K2.6 moonshotai | author-reported Author reports mmmu_pro_vision 79.4decision-model estimate p=0.97 +1 | 1.03T | 256K | $0.45 | Cluster | Kimi (Modified MIT) (modi… gated-threshold | 5 | — |
| Kimi-K2.5 moonshotai | author-reported Author reports MMMU-Pro 78.5Author reports MathVision 84.2 +1 | 1.03T | 256K | $0.45 | Cluster | Kimi (Modified MIT) (modi… gated-threshold | 4 | — |
| gemma-4-E4B-it google | author-reported Author reports MMMU Pro 52.6%Author reports MATH-Vision 59.5% +1 | 8.00B | 128K | — | Single GPU | Apache 2.0 | 1 | — |
| Qwen2.5-VL-7B-Instruct Qwen | author-reported Author reports MMStar 63.9Author reports MathVision 25.07 +1 | 8.29B | 125K | — | Single GPU | Apache 2.0 | 1 | — |
| Qwen2.5-VL-72B-Instruct Qwen | author-reported Author reports MMMU_Pro 51.1Author reports MathVista_MINI 74.8 +1 | 73.4B | 125K | $0.80 | Multi-GPU node | Qwen License (qwen) gated-threshold | 3 | — |
| Muse-Glimmer-30B meta-models | author-reported Author reports MMMU Pro 74decision-model estimate p=1.00 +1 | 29.8B | 128K | $0.30 | Single high-end GPU | Apache 2.0 | 4 | — |
| gemma-4-E2B-it google | author-reported Author reports MMMU Pro 44.2%Author reports MATH-Vision 52.4% +1 | 5.12B | 128K | — | Single GPU | Apache 2.0 | 0 | — |
| gemma-4-12B-it google | author-reported Author reports MMMU Pro 69.1%Author reports MATH-Vision 79.7% +1 | 12.0B | 256K | — | Single high-end GPU | Apache 2.0 | 0 | — |
| Qwen2.5-VL-32B-Instruct Qwen | author-reported Author reports MMMU 70Author reports MMMU Pro 49.5 +1 | 33.5B | 125K | — | Multi-GPU node | Apache 2.0 | 2 | — |
| Inkling thinkingmachines | author-reported Author reports MMMU Pro (Standard 10) 73.5%Author reports mmmu_pro_standard_10_options 73.5 +1 | 952B | 1M | $0.00 | Cluster | Apache 2.0 | 4 | — |
| Qwen2-VL-2B-Instruct Qwen | author-reported Author reports RealWorldQA 62.9Author reports MMStar 48.0 +1 | 2.21B | 32K | — | Single GPU | Apache 2.0 | 0 | — |
| Qwen2-VL-7B-Instruct Qwen | author-reported Author reports RealWorldQA 70.1Author reports MMStar 60.7 +1 | 8.29B | 32K | — | Single GPU | Apache 2.0 | 0 | — |
| Florence-2-large microsoft | author-reported Author reports Refcocog val Accuracy 68.0decision-model estimate p=1.00 +1 | 777M | 4K | — | Laptop / edge | MIT | 0 | — |
| diffusiongemma-26B-A4B-it google | author-reported Author reports MMMU Pro 54.3%Author reports MATH-Vision 70.5% +1 | 25.8B (4B active) | 256K | — | Single high-end GPU | Apache 2.0 | 0 | — |
| gemma-4-31B google | author-reported Author reports MMMU Pro 76.9%Author reports MATH-Vision 85.6% +1 | 31.3B | 256K | $0.00 | Single high-end GPU | Apache 2.0 | 1 | — |
| gemma-4-E4B google | author-reported Author reports MMMU Pro 52.6%Author reports MATH-Vision 59.5% +1 | 8.00B | 128K | — | Single GPU | Apache 2.0 | 0 | — |
| Florence-2-base microsoft | author-reported Author reports Refcocog val Accuracy 66.3Author reports VQAv2 test-dev Acc 79.7 +1 | 232M | 1K | — | Laptop / edge | MIT | 0 | — |
| gemma-4-26B-A4B google | author-reported Author reports MMMU Pro 73.8%Author reports MATH-Vision 82.4% +1 | 25.8B (4B active) | 256K | $0.00 | Single high-end GPU | Apache 2.0 | 2 | — |
| gemma-4-E2B google | author-reported Author reports MMMU Pro 44.2%Author reports MATH-Vision 52.4% +1 | 5.12B | 128K | — | Single GPU | Apache 2.0 | 0 | — |
| gemma-4-12B google | author-reported Author reports MMMU Pro 69.1%Author reports MATH-Vision 79.7% +1 | 12.0B | 256K | — | Single high-end GPU | Apache 2.0 | 0 | — |
| Qianfan-OCR baidu | author-reported Author reports ChartQA 88.1Author reports TextVQA 80.0 +1 | 4.74B | 32K | — | Single GPU | Apache 2.0 | 0 | — |
| gemma-4-31B-it-assistant google | author-reported Author reports MMMU Pro 76.9%Author reports MATH-Vision 85.6% +1 | 470M | 256K | — | Laptop / edge | Apache 2.0 | 0 | — |
| Inkling-Small thinkingmachines | author-reported Author reports mmmu_pro_standard_10_options 74decision-model estimate p=1.00 +1 | 266B | 1M | $0.00 | Multi-GPU node | Apache 2.0 | 2 | — |
| Kimi-VL-A3B-Instruct moonshotai | author-reported Author reports MMMU-Val (Pass@1) 61.7Author reports MMBench-EN-v1.1 (Acc) 83.1 +1 | 16.4B (3B active) | 128K | — | Single high-end GPU | MIT | 0 | — |
| gemma-4-26B-A4B-it-assistant google | author-reported Author reports MMMU Pro 73.8%Author reports MATH-Vision 82.4% +1 | 420M (4B active) | 256K | — | Laptop / edge | Apache 2.0 | 0 | — |
| QVQ-72B-Preview Qwen | author-reported Author reports MMMU(val) 70.3Author reports MathVista(mini) 71.4 +1 | 73.4B | 125K | — | Multi-GPU node | Qwen License (qwen) gated-threshold | 1 | — |
| Llama-3.1-Nemotron-Nano-VL-8B-V1 nvidia | author-reported Author reports MMMU Val with chatGPT as a judge 48.2%Author reports AI2D 85.0% +1 | 8.72B | — | — | Single GPU | NVIDIA Open Model License… attribution | 0 | — |
| SmolVLM2-2.2B-Instruct HuggingFaceTB | author-reported Author reports Mathvista 51.5Author reports MMMU 42 +1 | 2.25B | 8K | — | Single GPU | Apache 2.0 | 0 | — |
| Qwen2-VL-72B-Instruct Qwen | author-reported Author reports RealWorldQA 77.8Author reports MMStar 68.3 +1 | 73.4B | 32K | — | Multi-GPU node | tongyi-qianwen review | 0 | — |
| Florence-2-base-ft microsoft | author-reported Author reports VQAv2 test-dev Acc 79.7Author reports TextVQA test-dev Acc 63.6 +1 | 232M | 1K | — | Laptop / edge | MIT | 0 | — |
| Florence-2-large-ft microsoft | author-reported Author reports VQAv2 test-dev Acc 81.7Author reports TextVQA test-dev Acc 73.5 +1 | 770M | 1K | — | Laptop / edge | MIT | 0 | — |
| NVLM-D-72B nvidia | author-reported Author reports MMMU (val / test) 58.7 / 54.9Author reports MathVista 65.2 +1 | 79.4B | — | — | Multi-GPU node | CC BY-NC 4.0 non-commercial | 0 | — |
| MiniCPM-V-4_5 openbmb | author-reported Author reports Video-MME 73.5decision-model estimate p=1.00 +1 | 8.70B | 40K | — | Single GPU | Apache 2.0 | 0 | — |
| Idefics3-8B-Llama3 HuggingFaceM4 | author-reported Author reports MMMU <br>(val) 46.6Author reports MathVista <br>(test) 58.4 +1 | 8.46B | 128K | — | Single GPU | Apache 2.0 | 0 | — |
| paligemma2-3b-pt-224 google | author-reported Author reports [AI2D][ai2d] 84.6decision-model estimate p=1.00 +1 | 3.03B | — | — | Single GPU | Gemma Terms of Use attribution | 0 | — |
| Mage-VL microsoft | author-reported Author reports AI2D w/ Mask 83.16Author reports ChartQA 84.88 +1 | 4.74B | 256K | — | Single GPU | Apache 2.0 | 0 | — |
| LFM2.5-VL-450M LiquidAI | author-reported Author reports MMStar 43.00Author reports RealWorldQA 58.43 +1 | 449M | 125K | — | Laptop / edge | lfm1.0 review | 0 | — |
| Kimi-VL-A3B-Thinking moonshotai | author-reported Author reports MathVision (full) (Pass@1) 36.8Author reports MathVista (mini) (Pass@1) 71.3 +1 | 16.4B (3B active) | 128K | — | Single high-end GPU | MIT | 0 | — |
| Step3-VL-10B stepfun-ai | author-reported Author reports MMMU 80.11Author reports MathVista 85.50 +1 | 10.2B | 64K | — | Single high-end GPU | Apache 2.0 | 0 | — |
| LFM2.5-VL-1.6B LiquidAI | author-reported Author reports MMStar 50.67Author reports BLINK 48.82 +1 | 1.60B | 125K | — | Laptop / edge | lfm1.0 review | 0 | — |
| EXAONE-4.5-33B LGAI-EXAONE | author-reported Author reports MMMU 78.7Author reports MMMU-Pro 68.6 +1 | 34.4B | 256K | — | Multi-GPU node | exaone review | 0 | — |
| Kimi-VL-A3B-Thinking-2506 moonshotai | author-reported Author reports MMBench-EN-v1.1 (Acc) 84.4Author reports RealWorldQA (Acc) 70.0 +1 | 16.4B (3B active) | 128K | — | Single high-end GPU | MIT | 0 | — |
| Eikos-27B-FP8 caiovicentino1 | author-reported Author reports MMStar 70.7%decision-model estimate p=0.99 +1 | 27.8B | 256K | — | Single high-end GPU | MIT | 0 | — |
| Qwen3.8-27B Qwen | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 27.8B | 256K | $0.02 | Single high-end GPU | Apache 2.0 | 7 | 28.6 s · 87 tok/s |
| GLM-5.3-Flash zai-org | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 321B | 1M | $0.02 | Cluster | MIT | 9 | 8.1 s · 153 tok/s |
| Kimi-K3 moonshotai | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 2.78T | 1M | $0.23 | Cluster | Kimi K3 (Modified MIT) (k… gated-threshold | 7 | 25.6 s · 59 tok/s |
| Qwen3.5-9B Qwen | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 9.65B | 256K | $0.08 | Single GPU | Apache 2.0 | 6 | 2.8 s · 117 tok/s |
| Qwen3.6-35B-A3B Qwen | structural decision-model estimate p=0.99tag: image-text-to-text +1 | 36.0B (3B active) | 256K | $0.05 | Multi-GPU node | Apache 2.0 | 5 | — |
| Qwen3-VL-8B-Instruct Qwen | structural decision-model estimate p=1.00name: VL +1 | 8.77B | 256K | $0.25 | Single GPU | Apache 2.0 | 4 | — |
| Qwen3.5-397B-A17B Qwen | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 403B (17B active) | 256K | $0.39 | Cluster | Apache 2.0 | 7 | — |
| gemma-3-4b-it google | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 4.30B | 128K | $0.04 | Single GPU | Gemma Terms of Use attribution | 4 | — |
| Qwen3.6-27B Qwen | structural decision-model estimate p=0.99tag: image-text-to-text +1 | 27.8B | 256K | $0.20 | Single high-end GPU | Apache 2.0 | 4 | — |
| Qwen3.5-35B-A3B Qwen | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 36.0B (3B active) | 256K | $0.07 | Multi-GPU node | Apache 2.0 | 4 | — |
| gemma-3-27b-it google | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 27.4B | 128K | $0.08 | Single high-end GPU | Gemma Terms of Use attribution | 4 | — |
| Qwen3.8-Flash-Next Qwen | structural decision-model estimate p=0.96tag: image-text-to-text +1 | 180B | 256K | $0.15 | Multi-GPU node | Qwen License (qwen-commun… gated-threshold | 2 | — |
| Qwen3.5-27B Qwen | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 27.8B | 256K | $0.20 | Single high-end GPU | Apache 2.0 | 4 | — |
| DeepSeek-V4.1-Flash-1017 deepseek-ai | structural decision-model estimate p=0.91tag: image-text-to-text +1 | 763B | 1M | $0.20 | Cluster | MIT | 4 | — |
| Qwen3.5-4B Qwen | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 4.66B | 256K | — | Single GPU | Apache 2.0 | 2 | — |
| Qwen3-VL-30B-A3B-Instruct Qwen | structural decision-model estimate p=1.00name: VL +1 | 31.1B (3B active) | 256K | $0.15 | Single high-end GPU | Apache 2.0 | 4 | — |
| DeepSeek-OCR deepseek-ai | structural decision-model estimate p=0.99tag: image-text-to-text +1 | 3.34B | 8K | — | Single GPU | MIT | 1 | — |
| Llama-4-Scout-17B-16E-Instruct meta-llama | structural decision-model estimate p=0.98tag: image-text-to-text +1 | 109B | 1.3M | $0.08 | Multi-GPU node | Llama Community License (… attribution | 4 | — |
| gemma-3-12b-it google | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 12.2B | 128K | $0.04 | Single high-end GPU | Gemma Terms of Use attribution | 3 | — |
| Unlimited-OCR baidu | structural decision-model estimate p=0.98tag: image-text-to-text +1 | 3.34B | 32K | — | Single GPU | MIT | 0 | — |
| gemma-3-1b-it google | structural decision-model estimate p=0.94 | 1000M | 32K | — | Laptop / edge | Gemma Terms of Use attribution | 1 | — |
| Qwen3-VL-4B-Instruct Qwen | structural decision-model estimate p=1.00name: VL +1 | 4.44B | 256K | — | Single GPU | Apache 2.0 | 2 | — |
| Qwen2.5-VL-3B-Instruct Qwen | structural decision-model estimate p=1.00name: VL +1 | 3.75B | 125K | — | Single GPU | Qwen License (qwen-resear… gated-threshold | 1 | — |
| Qwen3-VL-2B-Instruct Qwen | structural decision-model estimate p=1.00name: VL +1 | 2.13B | 256K | — | Single GPU | Apache 2.0 | 1 | — |
| Qwen3.5-122B-A10B Qwen | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 125B (10B active) | 256K | $0.26 | Multi-GPU node | Apache 2.0 | 3 | — |
| Qwen3.5-2B Qwen | structural decision-model estimate p=0.99tag: image-text-to-text +1 | 2.27B | 256K | — | Single GPU | Apache 2.0 | 2 | — |
| Qwen3-VL-235B-A22B-Instruct Qwen | structural decision-model estimate p=1.00name: VL +1 | 236B (22B active) | 256K | $0.20 | Multi-GPU node | Apache 2.0 | 3 | — |
| Qwen3.5-0.8B Qwen | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 873M | 256K | — | Laptop / edge | Apache 2.0 | 1 | — |
| Kimi-K2.7-Code moonshotai | structural decision-model estimate p=0.77tag: image-text-to-text +1 | 1.03T | 256K | $0.67 | Cluster | Kimi (Modified MIT) (modi… gated-threshold | 4 | — |
| Qwen3-VL-Embedding-2B Qwen | structural name: VLtag: image-text-to-text +1 | 2.13B | 256K | — | Single GPU | Apache 2.0 | 1 | — |
| UI-TARS-1.5-7B ByteDance-Seed | structural decision-model estimate p=0.99tag: multimodal +1 | 8.29B | 125K | $0.10 | Single GPU | Apache 2.0 | 3 | — |
| Llama-3.2-11B-Vision-Instruct meta-llama | structural decision-model estimate p=1.00name: Vision +1 | 10.7B | — | — | Single high-end GPU | Llama 3.2 Community Licen… attribution | 1 | — |
| DeepSeek-OCR-2 deepseek-ai | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 3.39B | 8K | — | Single GPU | Apache 2.0 | 0 | — |
| DeepSeek-V4-Flash-Vision-Exp deepseek-ai | structural decision-model estimate p=1.00name: Vision +1 | 305B | 1M | $0.22 | Cluster | MIT | 2 | — |
| Qwen2.5-Omni-7B Qwen | structural decision-model estimate p=1.00name: Omni +1 | 10.7B | — | — | Single high-end GPU | Apache 2.0 (apache-2.0) | 0 | — |
| Qwen3-VL-32B-Instruct Qwen | structural decision-model estimate p=1.00name: VL +1 | 33.4B | 256K | — | Multi-GPU node | Apache 2.0 | 2 | — |
| medgemma-4b-it google | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 4.30B | — | — | Single GPU | health-ai-developer-found… review | 0 | — |
| Qwen3-VL-Embedding-8B Qwen | structural decision-model estimate p=0.74name: VL +1 | 8.14B | 256K | — | Single GPU | Apache 2.0 | 0 | — |
| Qwen3-VL-235B-A22B-Thinking Qwen | structural decision-model estimate p=1.00name: VL +1 | 236B (22B active) | 256K | — | Multi-GPU node | Apache 2.0 | 3 | — |
| GLM-4.5V zai-org | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 108B | 64K | $0.60 | Multi-GPU node | MIT | 3 | — |
| Qwen3-Omni-30B-A3B-Instruct Qwen | structural decision-model estimate p=1.00name: Omni +1 | 35.3B (3B active) | — | — | Multi-GPU node | Apache 2.0 (apache-2.0) | 0 | — |
| gemma-3-270m google | structural decision-model estimate p=0.94 | 268M | 32K | — | Laptop / edge | Gemma Terms of Use attribution | 0 | — |
| HunyuanOCR tencent | structural decision-model estimate p=0.99tag: image-text-to-text +1 | 1.12B | 128K | — | Laptop / edge | tencent-hunyuan-community review | 0 | — |
| LocateAnything-3B nvidia | structural decision-model estimate p=1.00tag: vision +1 | 3.83B | 32K | — | Single GPU | NVIDIA Open Model License… attribution | 0 | — |
| Phi-3.5-vision-instruct microsoft | structural decision-model estimate p=1.00name: vision +1 | 4.15B | 128K | — | Single GPU | MIT | 0 | — |
| GOT-OCR2_0 stepfun-ai | structural decision-model estimate p=0.99tag: image-text-to-text +1 | 716M | 32K | — | Laptop / edge | Apache 2.0 | 0 | — |
| embeddinggemma-2 google | structural tag: multimodal | 744M | — | — | Laptop / edge | Apache 2.0 | 0 | — |
| Qwen3-VL-8B-Thinking Qwen | structural decision-model estimate p=1.00name: VL +1 | 8.77B | 256K | — | Single GPU | Apache 2.0 | 3 | — |
| Janus-Pro-7B deepseek-ai | structural decision-model estimate p=1.00pipeline: any-to-any +1 | — | — | — | — | MIT (deepseek) | 1 | — |
| medgemma-1.5-4b-it google | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 4.30B | — | — | Single GPU | health-ai-developer-found… review | 0 | — |
| LightOnOCR-2-1B lightonai | structural decision-model estimate p=0.98tag: image-text-to-text +1 | 1.01B | 16K | — | Laptop / edge | Apache 2.0 | 1 | — |
| GLM-4.6V-Flash zai-org | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 10.3B | 128K | $0.30 | Single high-end GPU | MIT | 2 | — |
| paligemma-3b-pt-224 google | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 2.92B | — | — | Single GPU | Gemma Terms of Use attribution | 0 | — |
| Cosmos-Reason2-2B nvidia | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 2.44B | — | — | Single GPU | NVIDIA Open Model License… attribution | 0 | — |
| gemma-3-270m-it google | structural decision-model estimate p=0.88 | 268M | 32K | — | Laptop / edge | Gemma Terms of Use attribution | 0 | — |
| Qwen2.5-Omni-3B Qwen | structural decision-model estimate p=1.00name: Omni +1 | 5.54B | — | — | Single GPU | Qwen License (qwen-resear… gated-threshold | 0 | — |
| moondream2 vikhyatk | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 1.93B | — | — | Single GPU | Apache 2.0 | 0 | — |
| granite-docling-258M ibm-granite | structural decision-model estimate p=0.99tag: image-text-to-text +1 | 258M | 8K | — | Laptop / edge | Apache 2.0 | 0 | — |
| GLM-4.1V-9B-Thinking zai-org | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 10.3B | 64K | — | Single high-end GPU | MIT | 0 | — |
| MiMo-V2.6-Pro-RL XiaomiMiMo | structural decision-model estimate p=1.00tag: multimodal +1 | 1.02T | 1M | $0.39 | Cluster | MIT | 2 | — |
| blip2-opt-2.7b Salesforce | structural decision-model estimate p=1.00tag: vision +1 | 3.74B | — | — | Single GPU | MIT | 0 | — |
| chandra-ocr-2 datalab-to | structural decision-model estimate p=0.88tag: image-text-to-text +1 | 5.30B | 256K | — | Single GPU | OpenRAIL attribution | 0 | — |
| deepseek-vl2-tiny deepseek-ai | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 3.37B | — | — | Single GPU | DeepSeek License (deepsee… attribution | 0 | — |
| Llama-4-Maverick-17B-128E-Instruct meta-llama | structural decision-model estimate p=0.96tag: image-text-to-text +1 | 402B | 512K | $0.19 | Cluster | Llama Community License (… attribution | 2 | — |
| MiMo-V2.6-Flash-RL XiaomiMiMo | structural decision-model estimate p=1.00tag: multimodal +1 | 311B | 1M | $0.09 | Cluster | MIT | 2 | — |
| Qwen-VL-Chat Qwen | structural decision-model estimate p=1.00name: VL +1 | — | 8K | — | — | Not declared review | 0 | — |
| Qwen3-VL-4B-Thinking Qwen | structural decision-model estimate p=1.00name: VL +1 | 4.44B | 256K | — | Single GPU | Apache 2.0 | 2 | — |
| surya-ocr-2 datalab-to | structural decision-model estimate p=0.91tag: image-text-to-text +1 | 686M | 256K | — | Laptop / edge | OpenRAIL attribution | 1 | — |
| Step-3.7-Flash stepfun-ai | structural decision-model estimate p=1.00tag: multimodal +1 | 201B | 256K | $0.20 | Multi-GPU node | Apache 2.0 | 3 | — |
| gemma-3-4b-pt google | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 4.30B | — | — | Single GPU | Gemma Terms of Use attribution | 1 | — |
| translategemma-4b-it google | structural tag: image-text-to-textpipeline: image-text-to-text +1 | 4.30B | 128K | — | Single GPU | Gemma Terms of Use attribution | 0 | — |
| Ornith-1.0-9B ornith-ai | structural tag: image-text-to-textconfig: vision_config | 1.47M | 256K | — | Laptop / edge | MIT | 1 | — |
| Llama-Guard-4-12B meta-llama | structural decision-model estimate p=0.95tag: image-text-to-text +1 | 12.0B | 160K | $0.14 | Single high-end GPU | Llama Community License (… attribution | 2 | — |
| Qwen3-VL-30B-A3B-Thinking Qwen | structural decision-model estimate p=1.00name: VL +1 | 31.1B (3B active) | 256K | $0.29 | Single high-end GPU | Apache 2.0 | 2 | — |
| GLM-4.6V zai-org | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 108B | 128K | $0.30 | Multi-GPU node | MIT | 3 | — |
| gemma-3n-E2B-it google | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 5.44B | 32K | — | Single GPU | Gemma Terms of Use attribution | 0 | — |
| MiniCPM-o-4_5 openbmb | structural decision-model estimate p=1.00tag: multimodal +1 | 9.37B | 40K | — | Single GPU | Apache 2.0 | 0 | — |
| MiMo-V2.5 XiaomiMiMo | structural decision-model estimate p=1.00tag: multimodal +1 | 311B | 1M | $0.12 | Cluster | MIT | 2 | — |
| medgemma-27b-it google | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 27.4B | — | — | Single high-end GPU | health-ai-developer-found… review | 0 | — |
| Janus-Pro-1B deepseek-ai | structural decision-model estimate p=1.00pipeline: any-to-any +1 | — | — | — | — | MIT (deepseek) | 1 | — |
| gemma-3n-E4B-it google | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 7.85B | 32K | — | Single GPU | Gemma Terms of Use attribution | 0 | — |
| Phi-3-vision-128k-instruct microsoft | structural decision-model estimate p=1.00name: vision +1 | 4.15B | 128K | — | Single GPU | MIT | 0 | — |
| MiniCPM-o-2_6 openbmb | structural decision-model estimate p=1.00tag: vision +1 | 8.67B | 32K | — | Single GPU | Apache 2.0 | 0 | — |
| Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled Jackrong | structural decision-model estimate p=0.85tag: image-text-to-text +1 | 27.8B | 256K | — | Single high-end GPU | Apache 2.0 | 1 | — |
| gemma-3-1b-pt google | structural decision-model estimate p=0.95 | 1000M | 32K | — | Laptop / edge | Gemma Terms of Use attribution | 1 | — |
| MiniMax-M3.1-Flash MiniMaxAI | structural tag: multimodalpipeline: image-text-to-text | 220B | 977K | $0.15 | Multi-GPU node | MiniMax Model License (mi… gated-threshold | 1 | — |
| dots.ocr dots-studio | structural decision-model estimate p=0.99tag: image-text-to-text +1 | 3.04B | 128K | — | Single GPU | MIT | 0 | — |
| Qwen3.5-0.8B-Base Qwen | structural decision-model estimate p=0.95tag: image-text-to-text +1 | 873M | 256K | — | Laptop / edge | Apache 2.0 | 0 | — |
| MiniCPM-V-4.6 openbmb | structural decision-model estimate p=1.00tag: multimodal +1 | 1.30B | 256K | — | Laptop / edge | Apache 2.0 | 0 | — |
| Ornith-1.0-35B ornith-ai | structural tag: image-text-to-textconfig: vision_config | 665K | 256K | — | Laptop / edge | MIT | 0 | — |
| OvisOCR2 ATH-MaaS | structural decision-model estimate p=0.96tag: multimodal +1 | 853M | 256K | — | Laptop / edge | Apache 2.0 | 1 | — |
| Molmo-7B-D-0924 allenai | structural decision-model estimate p=1.00tag: multimodal +1 | 8.02B | 4K | — | Single GPU | Apache 2.0 | 0 | — |
| Qwen3-Omni-30B-A3B-Thinking Qwen | structural decision-model estimate p=1.00name: Omni +1 | 31.7B (3B active) | — | — | Single high-end GPU | Apache 2.0 (apache-2.0) | 0 | — |
| Qwen3.5-4B-Base Qwen | structural decision-model estimate p=0.95tag: image-text-to-text +1 | 4.66B | 256K | — | Single GPU | Apache 2.0 | 0 | — |
| paligemma-3b-mix-224 google | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 2.92B | — | — | Single GPU | Gemma Terms of Use attribution | 0 | — |
| d1-3B LiquidAI | structural decision-model estimate p=1.00tag: multimodal +1 | 3.12B | — | — | Single GPU | lfm1.0 review | 0 | — |
| Hunyuan-A13B-Instruct tencent | structural config: image_token_id | 80.4B (13B active) | 32K | $0.14 | Multi-GPU node | tencent-hunyuan-a13b review | 1 | — |
| Cosmos-Reason2-8B nvidia | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 8.77B | — | — | Single GPU | NVIDIA Open Model License… attribution | 0 | — |
| llava-1.5-7b-hf llava-hf | structural decision-model estimate p=1.00name: llava +1 | 7.06B | 4K | — | Single GPU | Llama 2 Community License attribution | 0 | — |
| idefics2-8b HuggingFaceM4 | structural decision-model estimate p=1.00tag: multimodal +1 | 8.40B | 32K | — | Single GPU | Apache 2.0 | 0 | — |
| gemma-3-12b-pt google | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 12.2B | — | — | Single high-end GPU | Gemma Terms of Use attribution | 1 | — |
| Llama-3.2-11B-Vision meta-llama | structural decision-model estimate p=1.00name: Vision +1 | 10.6B | — | — | Single high-end GPU | Llama 3.2 Community Licen… attribution | 0 | — |
| Qwen3.5-9B-Base Qwen | structural decision-model estimate p=0.96tag: image-text-to-text +1 | 9.65B | 256K | — | Single GPU | Apache 2.0 | 0 | — |
| NVIDIA-Nemotron-Parse-v1.1 nvidia | structural decision-model estimate p=0.98tag: image-text-to-text +1 | 957M | 9K | — | Laptop / edge | NVIDIA Open Model License… attribution | 0 | — |
| Qwen-AgentWorld-35B-A3B Qwen | structural tag: image-text-to-textconfig: vision_config | 34.7B (3B active) | 256K | — | Multi-GPU node | Apache 2.0 | 0 | — |
| llava-v1.5-7b liuhaotian | structural decision-model estimate p=1.00name: llava +1 | — | 4K | — | — | Not declared review | 0 | — |
| olmOCR-7B-0225-preview allenai | structural decision-model estimate p=0.98tag: image-text-to-text +1 | 8.29B | 32K | — | Single GPU | Apache 2.0 | 0 | — |
| llava-v1.6-mistral-7b-hf llava-hf | structural decision-model estimate p=1.00name: llava +1 | 7.57B | 32K | — | Single GPU | Apache 2.0 | 0 | — |
| Qwen3.5-2B-Base Qwen | structural decision-model estimate p=0.95tag: image-text-to-text +1 | 2.27B | 256K | — | Single GPU | Apache 2.0 | 0 | — |
| Huihui-Qwen3.8-27B-abliterated huihui-ai | structural decision-model estimate p=0.99tag: image-text-to-text +1 | 27.8B | — | — | Single high-end GPU | Apache 2.0 | 1 | — |
| Qwen-VL Qwen | structural decision-model estimate p=0.92name: VL | — | — | — | — | Not declared review | 0 | — |
| InternVL3-78B OpenGVLab | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 78.4B | — | — | Multi-GPU node | Qwen License (qwen) gated-threshold | 0 | — |
| t5gemma-2-270m-270m google | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 786M | — | — | Laptop / edge | Gemma Terms of Use attribution | 0 | — |
| Ornith-1.5-9B ornith-ai | structural decision-model estimate p=0.77tag: image-text-to-text +1 | 9.65B | 256K | — | Single GPU | MIT | 1 | — |
| gemma-3n-E2B-it-litert-lm google | structural decision-model estimate p=0.99 | — | — | — | — | Gemma Terms of Use attribution | 0 | — |
| Qwen3-VL-2B-Thinking Qwen | structural decision-model estimate p=1.00name: VL +1 | 2.13B | 256K | — | Single GPU | Apache 2.0 | 0 | — |
| Llama-3.2-90B-Vision-Instruct meta-llama | structural decision-model estimate p=1.00name: Vision +1 | 88.6B | — | — | Multi-GPU node | Llama 3.2 Community Licen… attribution | 1 | — |
| Ornith-1.5-35B-A3B ornith-ai | structural decision-model estimate p=0.81tag: image-text-to-text +1 | 36.0B (3B active) | 256K | — | Multi-GPU node | MIT | 0 | — |
| translategemma-12b-it google | structural tag: image-text-to-textpipeline: image-text-to-text +1 | 12.2B | 128K | — | Single high-end GPU | Gemma Terms of Use attribution | 0 | — |
| MiniMax-VL-01 MiniMaxAI | structural decision-model estimate p=1.00name: VL +1 | 456B | 8K | — | Cluster | Not declared review | 0 | — |
| olmOCR-2-7B-1025 allenai | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 8.29B | 125K | — | Single GPU | Apache 2.0 | 1 | — |
| MiniCPM-Llama3-V-2_5 openbmb | structural decision-model estimate p=1.00tag: vision +1 | 8.54B | 8K | — | Single GPU | Not declared review | 1 | — |
| Fara-7B microsoft | structural decision-model estimate p=0.94tag: multimodal +1 | 8.29B | 125K | — | Single GPU | MIT | 0 | — |
| Janus-1.3B deepseek-ai | structural decision-model estimate p=1.00pipeline: any-to-any +1 | 2.09B | — | — | Single GPU | MIT (deepseek) | 0 | — |
| InternVL3-8B OpenGVLab | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 7.94B | — | — | Single GPU | Apache 2.0 (qwen) | 0 | — |
| gemma-3n-E4B-it-litert-lm google | structural decision-model estimate p=0.99 | — | — | — | — | Gemma Terms of Use attribution | 0 | — |
| InternVL3-1B OpenGVLab | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 938M | — | — | Laptop / edge | Apache 2.0 (qwen) | 0 | — |
| PaddleOCR-VL-1.6 PaddlePaddle | structural decision-model estimate p=1.00name: VL +1 | 959M | 128K | — | Laptop / edge | Apache 2.0 | 0 | — |
| granite-vision-4.1-4b ibm-granite | structural decision-model estimate p=1.00name: vision +1 | 4.00B | — | — | Single GPU | Apache 2.0 | 0 | — |
| Cosmos-Reason1-7B nvidia | structural decision-model estimate p=1.00tag: image-text-to-text +1 | 8.29B | 125K | — | Single GPU | NVIDIA Open Model License… attribution | 0 | — |
| gemma-4-E4B-it-assistant google | structural pipeline: any-to-any | 78.8M | — | — | Laptop / edge | Apache 2.0 | 0 | — |
| deepseek-vl-7b-chat deepseek-ai | structural decision-model estimate p=1.00name: vl +1 | 7.34B | — | — | Single GPU | DeepSeek License (deepsee… attribution | 0 | — |
| Qwen3.5-35B-A3B-Base Qwen | structural decision-model estimate p=0.89tag: image-text-to-text +1 | 36.0B (3B active) | — | — | Multi-GPU node | Apache 2.0 | 0 | — |
| GOT-OCR-2.0-hf stepfun-ai | structural decision-model estimate p=0.98tag: image-text-to-text +1 | 561M | — | — | Laptop / edge | Apache 2.0 | 0 | — |
| Intern-Decision-4B internlm | structural decision-model estimate p=0.82tag: multimodal +1 | 4.54B | 256K | — | Single GPU | Apache 2.0 | 0 | — |
| Intern-Decision-0.8B internlm | structural decision-model estimate p=0.77tag: multimodal +1 | 853M | 256K | — | Laptop / edge | Apache 2.0 | 0 | — |
| Winnow-12B EldanRing | structural decision-model estimate p=0.98tag: vision +1 | — | 256K | — | — | Apache 2.0 | 0 | — |
| Intern-Decision-2B internlm | structural decision-model estimate p=0.75tag: multimodal +1 | 2.21B | 256K | — | Single GPU | Apache 2.0 | 0 | — |
| Tev1-4B-experimental togethercomputer | structural tag: image-text-to-textconfig: vision_config | 4.66B | 256K | $0.04 | Single GPU | Not declared review | 1 | — |
| Winnow-E4B EldanRing | structural decision-model estimate p=0.93tag: image-text-to-text +1 | — | — | — | — | Apache 2.0 | 0 | — |
| Tev1-0.8B-experimental togethercomputer | structural tag: image-text-to-textconfig: vision_config | 873M | 256K | — | Laptop / edge | Not declared review | 0 | — |
| xor juspay | structural decision-model estimate p=0.99tag: image-text-to-text +1 | 35.1B | 256K | — | Multi-GPU node | Apache 2.0 | 0 | — |
| openjev openjev | structural decision-model estimate p=0.85config: vision_config | 27.4B | 256K | — | Single high-end GPU | CC BY-NC 4.0 non-commercial | 0 | — |
| Jev-Omni akhilaaa3 | structural decision-model estimate p=0.96name: Omni +1 | 12.0B | 256K | — | Single high-end GPU | Apache 2.0 | 0 | — |
| pplx-decider-v1-27b perplexity-ai | structural tag: multimodalconfig: vision_config | 26.1B | 256K | — | Single high-end GPU | Apache 2.0 | 0 | — |
| decider-2b Mapika | structural decision-model estimate p=0.99tag: vision | 1.88B | 256K | — | Single GPU | Apache 2.0 | 0 | — |
| decider-4b Mapika | structural decision-model estimate p=0.99tag: vision | 4.21B | 256K | — | Single GPU | Apache 2.0 | 0 | — |
| jpt-4b kirp | structural decision-model estimate p=0.79tag: multimodal +1 | 4.54B | 256K | — | Single GPU | CC BY-NC 4.0 non-commercial | 0 | — |
| metask-jev-4b-policy-mix wayfind | structural tag: image-text-to-textconfig: vision_config | 4.54B | 256K | — | Single GPU | Apache 2.0 | 0 | — |
| rune-26b-a4b-GGUF surogate | structural tag: vision | 25.8B | — | — | Single high-end GPU | Apache 2.0 | 0 | — |
| mojev MoLeMo-Lab | structural tag: multimodal | 854M | — | — | Laptop / edge | MIT | 0 | — |
| NeoHorse-Jev-4B TokenRhythm | structural tag: multimodal | — | — | — | — | Apache 2.0 | 0 | — |
| jpt-0.8b kirp | structural tag: multimodalpipeline: image-text-to-text +1 | 853M | 256K | — | Laptop / edge | CC BY-NC 4.0 non-commercial | 0 | — |
| jpt-9b kirp | structural tag: multimodalpipeline: image-text-to-text +1 | 9.41B | 256K | — | Single GPU | CC BY-NC 4.0 non-commercial | 0 | — |
| jebadiah-27b frontier-infra | structural tag: image-text-to-textpipeline: image-text-to-text +1 | 27.8B | 256K | — | Single high-end GPU | Apache 2.0 | 0 | — |
No model matches every constraint. Loosen one, or switch the evidence filter back to any.
How to read this
- measured a third-party leaderboard row for this model. Settings (agent harness, edit format, prompt) differ between boards; compare within a benchmark, not across.
- author-reported a number the model author published in the card or eval files. Not reproduced by us.
- structural a deterministic signal from the name, config or chat template, or a decision-model estimate. A weak signal: it says the model was built or tagged for the job, not how well it does it. "decision-model estimate" chips are a probability from the Jev decision model, not a measurement.
- Runs on is the memory class of the BF16 weights (method); choosing a class shows that class and smaller. Quantized builds may fit a class lower.
- Price from is the cheapest observed hosted input price, USD per million tokens; models with no observed price are hidden when a ceiling is set. License friction: permissive = none or attribution-only; high = non-commercial.
- Ordering within a tier is the watch-list score (Hub likes, downloads, recency), not a quality verdict.
Subscribe
Follow "Understand images and video"
- /feeds/for/vision-language.xml(RSS 2.0) · one item per model with measured or author-reported evidence for this job
- /feeds/for/vision-language.json(JSON Feed 1.1) · same items, JSON Feed
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.