Alternatives to Liquid d1-3B

Realtime-Venus is an Apache 2.0 9B full-duplex audio-visual model that can speak, interrupt, and delegate while it still listens.

XHToken's efficient 4B open-weight model: hybrid sliding-window attention for native 1M-token context, strong coding and agent performance, Apache 2.0, multilingual.

Mistral Large 4 is Mistral's public-preview multimodal MoE model with a 1M-token context window and open weights planned for release.

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

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

DeepSeek-V4.1-Flash is a MIT 552B multimodal MoE with 8B/16B activation, 1M context, and about 1/4 the KV cache of V4-Flash.

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

mini-AGI is a byte-level model that trains from scratch on an 8GB GPU and keeps learning from one stream without forgetting what it knew.

Mistral Shieldstral 1.0: live compact multimodal moderation model. Apache 2.0.

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

Ternary Bonsai 2 is a 27B Qwen3.8 ternary GGUF that the model card says keeps 98.2% of FP16 scores in a 5.95 GB pack.

Previous DeepSeek V2.5 snapshot. Not a 2026 chat+coder default.

Bilibili's open multilingual translation family: 150-language text, speech subtitles, controlled dubbing, and long-document translation built on Qwen3.5.

Moonshot AI's open-weight 2.8T multimodal agentic model with 1M-token context, the world's first open 3T-class model rivaling closed frontier models.