Qwen3.8-2.4T-A95B
PulseAugur coverage of Qwen3.8-2.4T-A95B — every cluster mentioning Qwen3.8-2.4T-A95B across labs, papers, and developer communities, ranked by signal.
- 2026-09-10 product_launch Alibaba released the open weights for its Qwen3.8-2.4T-A95B model. source
- 2026-09-09 product_launch Alibaba's Qwen team released Qwen3.8-2.4T-A95B, a 2.4 trillion parameter open-weight model. source
- 2026-09-09 product_launch Alibaba's Qwen team released Qwen3.8-2.4T-A95B, the first open-weights Qwen-Max-class model. source
- 2026-08-18 product_launch Alibaba Group released the open-weight checkpoint Qwen3.8-2.4T-A95B for its Qwen3.8-Max service on August 12, 2026. source
- 2026-08-14 product_launch Alibaba's Qwen team released the Qwen3.8-2.4T-A95B model, now available on SiliconFlow. source
- 2026-08-14 product_launch Alibaba's Qwen team launched the Qwen3.8-2.4T-A95B model, making it available on multiple inference platforms. source
- 2026-08-14 product_launch Alibaba has released Qwen3.8-2.4T-A95B, an open-weight model in its Qwen-Max class. source
- 2026-08-13 product_launch Alibaba released the Qwen3.8-2.4T-A95B, a large-scale MoE model, and the Flag OS community achieved multi-chip adaptation for it. source
- 2026-08-13 product_launch Alibaba's Qwen team released the 2.4 trillion parameter AI model Qwen3.8-2.4T-A95B as an open-source model. source
- 2026-08-12 product_launch Together AI has launched the Qwen3.8-2.4T-A95B model on its Serverless Inference platform. source
- 2026-08-12 product_launch The Qwen3.8-2.4T-A95B model has been launched on the Fireworks AI platform. source
- 2026-08-12 product_launch Alibaba's Qwen team released the Qwen3.8-2.4T-A95B large language model. source
- 2026-08-08 product_launch Qwen has released the Qwen3.8-2.4T-A95B model, with variants and support across multiple platforms. source
3 day(s) with sentiment data
Qwen3.8-2.4T-A95B leverages Gated DeltaNet and standard attention
The Qwen3.8-Max model utilizes a hybrid architecture combining Gated DeltaNet with standard attention mechanisms. This suggests a novel approach to managing parameter activation and computational efficiency in large MoE models.
Flag OS enables broad hardware compatibility for Qwen3.8-2.4T-A95B
The successful Day 0 multi-chip adaptation of Qwen3.8-2.4T-A95B across nine different AI chip architectures via Flag OS highlights the platform's effectiveness. This indicates a significant reduction in the time-to-deployment for large models on diverse hardware.
Qwen3.8-2.4T-A95B will see rapid adoption for long-context tasks
Given its 1 million token context window and strong performance in benchmarks like PaperBench, Qwen3.8-Max is likely to be quickly adopted by developers for applications requiring extensive context processing, such as complex coding and agentic systems.
-
Alibaba releases open weights for Qwen3.8-2.4T-A95B MoE model
Alibaba has released the open weights for its Qwen3.8-2.4T-A95B model. This Mixture-of-Experts (MoE) model boasts 2.4 trillion total parameters, with 95 billion active parameters. It features hybrid attention mechanisms…
-
Alibaba releases Qwen3.8-2.4T-A95B open-weight model for complex AI tasks
Alibaba's Qwen team has released Qwen3.8-2.4T-A95B, the first open-weight model in their Qwen-Max class. This large model, featuring a hybrid attention architecture and 2.4 trillion parameters with 95 billion activated …
-
August 2026: AI Models Focus on Efficiency and Specialization · 1 source tracked
August 2026 saw a surge of new AI models, with a particular focus on efficiency and specialized capabilities. Many releases emphasized sparse architectures, enabling faster inference and lower active parameter counts. S…
-
NVIDIA releases quantized DeepSeek and Qwen LLMs for Blackwell hardware
NVIDIA has released quantized versions of two large language models, DeepSeek-V4-Pro-0813 and Qwen3.8-2.4T-A95B, utilizing their NVFP4 quantization method. The DeepSeek model, with 1.65 trillion parameters, employs Hybr…
-
Alibaba releases open-weight Qwen3.8-2.4T-A95B checkpoint
Alibaba Group has released the open-weight checkpoint for its Qwen3.8-Max service, named Qwen3.8-2.4T-A95B. This release occurred on August 12, 2026, making the underlying model weights available. The information was re…
-
Alibaba's Qwen releases 2.4T parameter model for agentic tasks
Alibaba's Qwen team has released Qwen3.8-2.4T-A95B, a large language model with 2.4 trillion parameters and 95 billion active parameters. This model is designed for agentic tasks, capable of coding, researching, plannin…
-
Alibaba's Qwen3.8-2.4T-A95B model launches across multiple platforms
Alibaba's Qwen team has launched its Qwen3.8-2.4T-A95B model, a large sparse Mixture-of-Experts model with 2.4 trillion total parameters and 95 billion active parameters. This model is now available through various plat…
-
Alibaba's Qwen3.8-Max model launches with 2.4T parameters, 1M context · 2 sources tracked
Alibaba's Qwen team has released Qwen3.8-Max, a large mixture-of-experts model with 2.4 trillion parameters and 95 billion active parameters. This new model is designed for demanding tasks such as autonomous agents, hea…
-
Alibaba releases open-weight Qwen3.8-Max with 2.4T parameters
Alibaba has released Qwen3.8-2.4T-A95B, marking the first open-weight release of a model in its Qwen-Max class. This new model boasts 2.4 trillion total parameters, with 95 billion active parameters per forward pass, ut…
-
OpenAI's Jalapeño chip claims efficiency gains; agent systems evolve
OpenAI has released benchmark details for its custom inference chip, Jalapeño, claiming significant improvements in efficiency and latency over NVIDIA's GB200 and GB300 systems. The chip reportedly offers better perform…
-
DeepSeek, Qwen model releases reveal pricing nuances and licensing details
A recent analysis of four new AI model releases highlights key details often missed in benchmark comparisons. DeepSeek's pricing for its V4 Pro model will increase significantly in August 2026, particularly impacting wo…
-
Alibaba releases Qwen3.8-2.4T model, Flag OS enables multi-chip adaptation
Alibaba has open-sourced its largest and most capable model to date, Qwen3.8-2.4T-A95B, featuring a massive 2.4 trillion parameter MoE architecture with 95 billion active parameters. This release is accompanied by the F…
-
Alibaba releases 2.4T parameter Qwen3.8 open-source AI model
Alibaba's AI research team, Qwen, has released "Qwen3.8-2.4T-A95B" as an open-source model. This model, with 2.4 trillion parameters, is based on the "Qwen3.8 Max" which has demonstrated performance comparable to leadin…
-
DeepSeek V4 Pro launches, Alibaba open-sources Qwen3.8 model · 2 sources tracked
DeepSeek has officially launched its V4 Pro model, enhancing its agent capabilities and aiming to compete with models like Anthropic's Claude Fable 5. The new model supports a 1 million token context window and offers a…
-
Together AI launches Qwen3.8-2.4T-A95B for coding and agentic tasks
Together AI has launched Qwen3.8-2.4T-A95B, a new flagship model from the Qwen team, on its Serverless Inference platform. This model is designed for coding and long-horizon agent workflows, boasting 2.4 trillion parame…
-
Alibaba releases Qwen3.8-2.4T-A95B, rivaling Fable 5 and GPT 5.6 "Sol"
Alibaba's Qwen team has released a new large language model, Qwen3.8-2.4T-A95B, which rivals top commercial models like Fable 5 and GPT 5.6 "Sol". The model's release was announced on Mastodon and reported by PC Watch.
-
Fireworks AI adds DeepSeek-V4-Pro and Qwen3.8 models
Fireworks AI has announced the availability of two new large language models on its platform: DeepSeek-V4-Pro-0813 and Qwen3.8-2.4T-A95B. DeepSeek-V4-Pro-0813 is designed for agentic and tool-augmented workflows, outper…
-
Alibaba releases Qwen3.8 models in 27B and 2.4T parameter sizes · 8 sources tracked
Alibaba's Qwen team has released several versions of its Qwen3.8 model, including the 27B parameter version and the 2.4T parameter version. These models are available on Hugging Face, with some noted as pre-release or F…
-
Alibaba releases Qwen3.8 open-weight models, gaining traction on leaderboards
Alibaba's Qwen has released its Qwen3.8 series of open-weight models, including Qwen3.8-27B and Qwen3.8-2.4T-A95B. The Qwen3.8-27B model boasts a 262K native context window, extendable to 1M tokens via YaRN, and has ach…
-
Alibaba's Qwen3.8-27B model released; AI aids GPU porting; LLM infra detailed
Alibaba's Qwen team has released Qwen3.8-27B, a dense 27-billion parameter model that fits on a single GPU and supports a 1 million token context window, with Day-0 integration in vLLM. Concurrently, research is explori…