Previous Qwen3 embedding snapshot.

Models, datasets, and inference. Extra private storage $18/TB. Not a 100k-model slogan.

Previous CogVLM snapshot. Not a 2026 VLM default.

Previous MythoMax 13B snapshot.

Previous OpenChat 3.5 7B snapshot.

Previous Qwen2-VL 72B Instruct snapshot. Not a 2024 VL crown.

Previous Rocinante 12B snapshot.

Previous LLaVA-NeXT / LLaVA 1.6 snapshot.

Previous Alibaba Qwen-VL snapshot. Not a 2026 LVLM default.

Previous Qwen3-VL embedding snapshot.

Previous Qwen3-VL rerank snapshot.

Previous DeepSeek-VL 7B Base snapshot. Not a 2026 VL default.

Agnes-3.0-Flash Preview is an Apache-2.0 33B multimodal open-weight checkpoint with 262k context, not the proprietary API model.

Apple's 9B vision-language model reads documents as compressed images, then expands only the pages that matter, holding accuracy to 4.3x compression.

DeepSeek-V4-Flash-Vision-Exp is the first V4 multimodal experiment: MIT vision weights on Flash, 305B params, screenshots plus text agents.

Inco AI's Apache 2.0 block-diffusion drafter for speculative decoding. Released 2026-08-18 for Qwen3.8-27B and Muse Glimmer.

Ling-3.0-flash-VL is inclusionAI's MIT multimodal MoE: 124B total, 5.5B active, with image and video.

inclusionAI MIT hybrid-linear MoE: 124B total, 5.1B active.

Ling-3.0-tiny is inclusionAI's MIT hybrid-linear MoE: 7.9B total, 1.3B active, for local agents.

OpenBMB MiniCPM5-2B is a dense 2B on-device LLM with 131k context, Apache-2.0 weights, and cookbooks for vLLM, SGLang, and llama.cpp.

Nex-N2.5-mini is Nex-AGI's Apache-2.0 multimodal agent model for computer use, browsing, and coding.

Nex-N2.5-Pro is Nex-AGI's Apache-2.0 multimodal agent checkpoint for computer use, browsing, and coding.

Orukeet is a 25-language ASR model that freezes fitted Gabor kernels on NVIDIA Parakeet TDT 0.6B v3.

Qwen-Drive-1.0-4B is Alibaba's Apache-2.0 driving VLM that unifies 3D perception, VQA, and motion planning.