mlsanityData to 7 Oct 2026

Zamba2-VL Technical Report

arXiv 2606.00390 ↗First posted 29 May 2026cs.CV cs.AIcited by 0

On the map: Multimodal LLMs / vision-language models → MLLM training and visual instruction tuning

Abstract (arXiv)

We present Zamba2-VL, a suite of vision-language models built on Zamba2, a hybrid language-model architecture combining Mamba2 state-space layers with a small number of shared transformer blocks. Across a broad range of image understanding, reasoning, OCR, grounding, and counting benchmarks, Zamba2-VL is competitive with leading Transformer-based open-weight VLMs of comparable scale, including the Molmo2, Qwen3-VL, and InternVL3.5 families, and substantially outperforms prior SSM-based and hybrid VLMs such as VL-Mamba, Cobra, and mmMamba. Inheriting the near-linear prefill compute and small, near-constant recurrent state of its Zamba2 backbone, Zamba2-VL delivers roughly an order of magnitude lower time-to-first-token (TTFT) than these Transformer baselines at matched parameter scale, with the efficiency gap most pronounced at the smaller 1.2B and 2.7B scales most relevant to on-device and edge deployment. We release three models -- 1.2B, 2.7B, and 7B -- together with inference code at https://huggingface.co/collections/Zyphra/zamba2-vl.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in MLLM training and visual instruction tuning, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 112 of its 135 references that are in the map. First posted on arXiv 29 May 2026.

Builds on

112 of 135 references.

Closest ancestors · 16

Same direction, newest first

  1. 2025Qwen3-VL Technical Report26 Nov
  2. 2025LLaVA-OneVision-1.5: Fully Open Framework for Democratized Multimodal Training28 Sep
  3. 2025Do You See Me : A Multidimensional Benchmark for Evaluating Visual Perception in Multimodal LLMs28 May
  4. 2025MaTVLM: Hybrid Mamba-Transformer for Efficient Vision-Language Modeling17 Mar
  5. 2025Multimodal Mamba: Decoder-only Multimodal State Space Model via Quadratic to Linear Distillation18 Feb
  6. 2024MAmmoTH-VL: Eliciting Multimodal Reasoning with Instruction Tuning at Scale6 Dec
  7. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution18 Sep
  8. 2024Shaking Up VLMs: Comparing Transformers and Structured State Space Models for Vision & Language Modeling9 Sep
  9. 2024LLaVA-OneVision: Easy Visual Task Transfer6 Aug
  10. 2024LLaVA-NeXT-Interleave: Tackling Multi-image, Video, and 3D in Large Multimodal Models10 Jul
  11. 2024BLINK: Multimodal Large Language Models Can See but Not Perceive18 Apr
  12. 2024Cobra: Extending Mamba to Multi-Modal Large Language Model for Efficient Inference21 Mar
4 more
  1. 2024VL-Mamba: Exploring State Space Models for Multimodal Learning20 Mar
  2. 2023ShareGPT4V: Improving Large Multi-Modal Models with Better Captions21 Nov
  3. 2023UReader: Universal OCR-free Visually-situated Language Understanding with Multimodal Large Language Model8 Oct
  4. 2023SEED-Bench: Benchmarking Multimodal LLMs with Generative Comprehension30 Jul

Other references · 96

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2024The Llama 3 Herd of Modelscited by 14,192
  3. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  4. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  5. 2023Segment Anythingcited by 8,158
  6. 2023Visual Instruction Tuningcited by 7,792
  7. 2023BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Modelscited by 6,069
  8. 2025Qwen2.5-VL Technical Reportcited by 5,418
  9. 2023Mamba: Linear-Time Sequence Modeling with Selective State Spacescited by 4,616
  10. 2021RoFormer: Enhanced Transformer with Rotary Position Embeddingcited by 4,581
  11. 2022Flamingo: a Visual Language Model for Few-Shot Learningcited by 4,337
  12. 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,131
38 more
  1. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,777
  2. 2023MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Modelscited by 2,689
  3. 2016Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answeringcited by 2,498
  4. 2023MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGIcited by 2,044
  5. 2022Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answeringcited by 1,881
  6. 2019Towards VQA Models That Can Readcited by 1,696
  7. 2023Evaluating Object Hallucination in Large Vision-Language Modelscited by 1,696
  8. 2025Gemma 3 Technical Reportcited by 1,686
  9. 2023MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contextscited by 1,582
  10. 2022ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoningcited by 1,449
  11. 2015Microsoft COCO Captions: Data Collection and Evaluation Servercited by 1,447
  12. 2020DocVQA: A Dataset for VQA on Document Imagescited by 1,288
  13. 2024Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Dualitycited by 1,283
  14. 2024OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environmentscited by 1,134
  15. 2025SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Featurescited by 1,127
  16. 2019OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledgecited by 1,114
  17. 2023MVBench: A Comprehensive Multi-modal Video Understanding Benchmarkcited by 1,086
  18. 2024Vision Mamba: Efficient Visual Representation Learning with Bidirectional State Space Modelcited by 1,051
  19. 2022Token Merging: Your ViT But Fastercited by 995
  20. 2024VMamba: Visual State Space Modelcited by 955
  21. 2015Generation and Comprehension of Unambiguous Object Descriptionscited by 897
  22. 2016A Diagram Is Worth A Dozen Imagescited by 793
  23. 2022A-OKVQA: A Benchmark for Visual Question Answering using World Knowledgecited by 702
  24. 2024SpatialVLM: Endowing Vision-Language Models with Spatial Reasoning Capabilitiescited by 679
  25. 2024PaliGemma: A versatile 3B VLM for transfercited by 679
  26. 2023HallusionBench: An Advanced Diagnostic Suite for Entangled Language Hallucination and Visual Illusion in Large Vision-Language Modelscited by 572
  27. 2023DriveLM: Driving with Graph Visual Question Answeringcited by 527
  28. 2024MMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmarkcited by 494
  29. 2023Gated Linear Attention Transformers with Hardware-Efficient Trainingcited by 490
  30. 2024Gated Delta Networks: Improving Mamba2 with Delta Rulecited by 483
  31. 2023OCRBench: On the Hidden Mystery of OCR in Large Multimodal Modelscited by 466
  32. 2024AndroidWorld: A Dynamic Benchmarking Environment for Autonomous Agentscited by 450
  33. 2024VLMEvalKit: An Open-Source Toolkit for Evaluating Large Multi-Modality Modelscited by 443
  34. 2024WebVoyager: Building an End-to-End Web Agent with Large Multimodal Modelscited by 424
  35. 2016Synthetic Data for Text Localisation in Natural Imagescited by 396
  36. 2024LLaVA-PruMerge: Adaptive Token Reduction for Efficient Large Multimodal Modelscited by 394
  37. 2024Jamba: A Hybrid Transformer-Mamba Language Modelcited by 390
  38. 2021Inter-GPS: Interpretable Geometry Problem Solving with Formal Language and Symbolic Reasoningcited by 376

50 most cited of 96 All 96 (JSON)

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Small Vision-Language Models: A Survey on Compact Architectures and Techniques
  2. 2024MobileVLM V2: Faster and Stronger Baseline for Vision Language Model
  3. 2025TransMamba: Fast Universal Architecture Adaption from Transformers to Mamba
  4. 2025Flash-VL 2B: Optimizing Vision-Language Model Performance for Ultra-Low Latency and High Throughput
  5. 2025TransMamba: A Sequence-Level Hybrid Transformer-Mamba Language Model
  6. 2026Empirical Recipes for Efficient and Compact Vision-Language Models
  7. 2024ML-Mamba: Efficient Multi-Modal Large Language Model Utilizing Mamba-2
  8. 2025MagicVL-2B: Empowering Vision-Language Models on Mobile Devices with Lightweight Visual Encoders via Curriculum Learning