open-weight

Unsloth has released GGUF quantizations of EmbeddingGemma 2, Google DeepMind's open 740M-parameter embedding model for text, images, video and audio. The base model maps all four modalities into one 768-dimensional space with an 8K context window under Apache 2.0.

Mistral AI has launched a public preview of Mistral Large 4, a 1-trillion-parameter natively multimodal model with 49 billion active parameters. The preview API is live on Mistral Studio, and open weights are promised by the end of October 2026. Pricing and context window have not been disclosed.

Reflection introduced Beam, a 501B-parameter mixture-of-experts model with 23B active parameters. Mistral said it is finishing Mistral Large 4, a 1T-parameter multimodal model with 49B active parameters. Both companies plan open-weight releases in October, and both are positioning the models against Chinese open-weight leaders.

Amazon Bedrock now offers GLM 5.3, Z.ai's 753B-parameter mixture-of-experts model, through managed APIs with cross-Region inference, prompt caching and service tiers. Access is limited to eligible enterprise customers. Pricing and context window were not disclosed in AWS's announcement.

Reflection AI has unveiled Beam, a text-only mixture-of-experts model with 501 billion total parameters, 23 billion active, and a 1 million token context window. The company claims it matches Z.ai's GLM-5.2 on advanced reasoning benchmarks while using 3-4x less inference compute. Weights and the full technical report are due later this month.

Axios reports, citing unnamed sources, that Nvidia-backed startup Reflection will soon release its first open-weight model. It is expected to lag the most advanced US models while competing with top Chinese open-weight systems. Parameter count, context window, pricing and benchmarks have not been disclosed.

Meta Superintelligence Labs has released Muse Glimmer 30B, a dense open-weight model distilled from its larger Muse Spark system and tuned for agentic workflows on consumer hardware. The model supports 131K context, image understanding, and over 100 languages at $0.30/$1.10 per 1M input/output tokens.

Z.ai, maker of the GLM model series, has confirmed it is behind Ox Alpha, the mysterious open-weight model that appeared anonymously on OpenRouter and topped benchmark leaderboards. The company will release the model's weights on Wednesday.

Alibaba's Qwen team has released Qwen3.8-Flash-Next, an open-weight model with 125B total parameters (6B activated) that previews architectural changes planned for Qwen4, including a new sparse attention mechanism and n-gram embeddings. The model natively supports 262,144 tokens of context, extensible to 1 million.

SenseNova has released SenseNova-U1.5-8B-MoT, an open-weight native multimodal model built on its NEO-unify architecture for image generation, editing, and native 4K output. The model is available on Hugging Face under an Apache 2.0 license, with no inference pricing yet since it must be self-hosted.

NVIDIA released the full-precision BF16 reference weights for Nemotron 3.5 Lightning, a 30B-parameter Mixture-of-Experts model with only 3B active parameters and support for up to 1 million tokens of context. The model uses a hybrid Mamba-2, MoE, and Attention architecture and is licensed under OpenMDW-1.1 for commercial use.

Unsloth has released GGUF quantizations of Kimi K3, a 2.8-trillion-parameter open-weight Mixture-of-Experts model from Moonshot AI with a 1-million-token context window and native vision support. The largest lossless quantization (Q8) weighs in at 1.56TB.

Black Forest Labs reports its FLUX.2 image generation models demonstrate more than 10 times fewer vulnerabilities for synthetic non-consensual intimate imagery (NCII) and child sexual abuse material (CSAM) compared to other leading open-weight models. The company claims targeted post-training mitigations reduced vulnerabilities by 77-98% before release, according to third-party red-teaming conducted by Cinder.

Chinese AI models from Tencent, Xiaomi, DeepSeek, MiniMax and Z.ai now occupy the top five positions by weekly token usage on OpenRouter, a major AI marketplace. The shift threatens to commoditize premium models from OpenAI and Anthropic as businesses reserve expensive frontier systems for only their hardest problems.

Moonshot AI's Kimi K3 model has matched or exceeded performance of Anthropic's Opus 4.8 and OpenAI's GPT-5.6 Sol in independent benchmarks while costing 40% less than comparable US models. The Beijing-based company plans to release Kimi K3 as an open-weight model on July 27.

Moonshot AI has released Kimi K3, an open-weight multimodal reasoning model with a 1-million token context window. The model is priced at $3 per 1M input tokens and $15 per 1M output tokens, available through OpenRouter.

Thinking Machines Lab released Inkling, a 975-billion-parameter mixture-of-experts model that uses 41 billion active parameters per task. The open-weight model was trained on 45 trillion tokens across text, image, audio, and video, marking the first public release from Mira Murati's AI startup.

Zhipu AI released its open-weight GLM-5.2 model, with researchers claiming it matches Anthropic's Mythos on certain bug-finding and cybersecurity tasks. The model lags behind Anthropic and OpenAI models on general benchmarks but represents a significant narrowing of capabilities between Chinese and US AI systems.

Ideogram has released Ideogram 4, its first open-weight text-to-image model with 9.3 billion parameters. The model supports native 2K resolution, structured JSON prompting with bounding-box layout controls, and is available in nf4 and fp8 quantizations under a non-commercial license.

Ideogram has released Ideogram 4, a 9.3B parameter open-weight text-to-image model trained from scratch. The model features structured JSON prompting, native 2K resolution output, and ranks as the top open-weight model on Design Arena. Available in fp8 and nf4 quantizations under a non-commercial license.

Chinese AI lab DeepSeek has released preview versions of DeepSeek V4 Flash and V4 Pro, mixture-of-experts models with 1 million token context windows. The V4 Pro has 1.6 trillion total parameters (49 billion active), making it the largest open-weight model available, while both models significantly undercut frontier model pricing.

Alibaba has released Qwen3.6-35B-A3B, the first open-weight model in the Qwen3.6 series. The model features 35B total parameters with 3B activated, a native 262K context window extensible to 1.01M tokens, and achieves 73.4% on SWE-bench Verified using 256 experts with 8 activated per token.

Zhipu AI has released GLM-5.1, a freely available open-weight model designed for long-running programming tasks that achieves 58.4% on SWE-Bench Pro, edging out GPT-5.4 (57.7%) and Claude Opus 4.6 (57.3%). The model's core capability is iterative strategy refinement—it rethinks its approach across hundreds of iterations and thousands of tool calls, recognizing dead ends and shifting tactics without human intervention. However, GLM-5.1 trails on reasoning and knowledge benchmarks, scoring 31% on Humanity's Last Exam compared to Gemini 3.1 Pro's 45%.

ByteDance has released Helios, a 14-billion-parameter open-weight video generation model that achieves 19.5 frames per second on a single GPU while generating minute-long video clips. The researchers claim this is the first model of its scale to reach near-real-time performance at this duration. Code and model weights are publicly available.

Alibaba has released Qwen3.5-35B-A3B, a 35-billion parameter multimodal model capable of processing images and text. The model is published under an Apache 2.0 license and available on Hugging Face with Transformers and SafeTensors format support.