Qwen3.8
Alibaba
时间线
已移出活跃追踪列表。尚未确认以此名称发布或更名。
- · 参考链接 · GitHub search · Alibaba · commits · onprem-ai/deployment-catalog · @tomas-polach
- · 参考链接 · GitHub search · Alibaba · commits · finereli/pyramid-mcp · @finereli
- · 参考链接 · GitHub search · Alibaba · commits · fdietze/pi-infinite-context · @fdietze
- · Alex 的文章 · 系列 · GitHub search · Alibaba · commits · agentgateway/agentgateway · @github-actions[bot]
摘录
json.providers.nebius.models["Qwen/Qwen3.8-27B"].name = "Qwen3.8 27B";
- · 参考链接 · GitHub search · Alibaba · commits · aabhay007/Personal-AI-Core · @aabhay007
- · 参考链接 · GitHub search · Alibaba · commits · QuantumKuba/Diffactly · @QuantumKuba
- · 参考链接 · GitHub search · Alibaba · commits · Hallo-ich-bin-PeppaPig/Singularity · @Hallo-ich-bin-PeppaPig
- · 参考链接 · GitHub search · Alibaba · commits · Azure/azureml-assets · @hanbitmyths
摘录
Add MAI, Qwen3.8 and Nemotron 3.5 Foundry Local coding models (#5452)
- · Alex 的文章 · 系列 · GitHub search · Alibaba · pulls · xadupre/devdash-agents · @xadupre
摘录
- Real Qwen3.8 27B on H200-101, physical CUDA 7: streaming generation and a follow-up turn succeeded. Used a temporary configuration with 16 KV blocks and
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] qwen38_flash package entry: register() and READMEs
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] MTP draft model
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] KV offload pool regions
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] Runner input adapter
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] Hyper-connection layers
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] GDN layer input preparation
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] Pooled KV cache materialization
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] PD transfer: identity, manifest and Raiden connector
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] Config hook, geometry checks and pooled cache layout
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] PLE feature: tables, host lookup, storage, offload and pooled PLE
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] Sequence-parallel prefill backbone
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] QSA attention layer
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] Model-owned GDN op and layer (no register_oot)
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] Per-group Mamba index accessors
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] Mamba tail fold: export prefill checkpoints from inside the step
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] In-place rows update op
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] PLE Torch-to-JAX ops
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] QSA Torch-to-JAX ops
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] Torch bridge for the QSA kernels
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] GDN input preparation kernel
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] Unified KV pool kernels
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] PLE lookup and gate kernels
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] Hyper-connection kernels
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] Pooled QSA kernel over the unified KV pool
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] QSA KV write kernels (fused, main, compressed)
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] QSA union-walk kernels (plan, dense, SparseCore gather, prefill)
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] Qwen3.8-Flash model and registration hooks
- · 参考链接 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
摘录
[Model][Qwen3.8] QSA main-KV decode context parallelism over rank pairs (DCP-2) kernels
- · 参考链接 · GitHub search · Alibaba · pulls · vjmichael/Lu-AI-GUI · @vjmichael
摘录
- Physical interactive owner Windows/Radeon Qwen3.8, Vulkan VRAM/TFFB TTFT p50/p95, and actual local installed MMProj; hosted CPU small models do not substitute.
- · 参考链接 · GitHub search · Alibaba · pulls · vitessio/vitess · @jdoupe
- · 参考链接 · GitHub search · Alibaba · pulls · nearai/cvm-compose-files · @PierreLeGuen
- · 参考链接 · GitHub search · Alibaba · pulls · michsindlinger/specwright · @michsindlinger
- · Alex 的文章 · 系列 · GitHub search · Alibaba · pulls · SceneWorks/SceneWorks · @michaeltrefry
摘录
SW-7119: Qwen3.8 27B describer model and tier in core, worker, API and catalog
- · Alex 的文章 · 系列 · GitHub search · Alibaba · pulls · NVIDIA/hyperflare-inference · @Raydius
摘录
The pinned Qwen3.8 Flash Next checkpoint trains a 262,144-token context (theta 10M, interleaved MRoPE 11/11/10). This adds a declared serving variant of the same weights
- · 参考链接 · GitHub search · Alibaba · pulls · FlashML-org/FreeToken · @zihaomu
- · Alex 的文章 · 系列 · GitHub search · Alibaba · pulls · BuQQzz/SHIELD2.0 · @BuQQzz
摘录
Makes Qwen3.8 27B (IQ1_M) a working coding model on an 8 GB card (RTX 2070 SUPER) through SHIELD's own llama-server.
- · 参考链接 · GitHub search · Alibaba · pulls · Azure/azureml-assets · @hanbitmyths
摘录
Add MAI, Qwen3.8 and Nemotron 3.5 Foundry Local coding models
- · 参考链接 · GitHub search · Alibaba · pulls · 0xburden/modelship · @0xburden
摘录
- Reviewed immutable catalog of 25 standalone Qwen3.8 GGUF quantizations; wizard displays sizes, numeric/letter/full-value selection and a visible UD-Q4_K_M default. No hardware-driven quant substitution; flags
- · 参考链接 · GitHub · QwenLM/qwen-code · PR
摘录
- The Standard preset has the same qwen3.8 effort and GLM thinking gaps and will be aligned in a follow-up
- · Alex 的文章 · 系列 · Alex Getman
摘录
🚨 Qwen3.8 Omni Flash is out It can see, hear and understand video, then actually act on it Give it hours of footage and it can find the right moments, translate, edit or turn it into notes. Give it a meeting and it can pull out tasks and start working on them 1M context, API from $0.43
- · Alex 的文章 · 系列 · Alex Getman
摘录
🚨 QWEN3.8 FLASH NEXT IS OUT The leaked specs were real: • 125B multimodal MoE • Only 6B active • 262K native context, up to 1M Qwen says it beats Claude Opus 4.6 across coding, agentic tasks, tool use and several multimodal benchmarks.
- · Alex 的文章 · 系列 · Alex Getman
摘录
🚨 QWEN3.8 FLASH NEXT DROPS TODAY A ModelScope description that was quickly taken down reportedly revealed: • 125B main parameters • 51B N gram embeddings • Only 6B active per token • Multimodal MoE architecture The model is also an official preview of the upcoming Qwen4 architecture.
- · Alex 的文章 · Alex Getman
摘录
⚡️ Qwen 3.8 27B beat GPT 5.6 Terra This happened in the AA Agentic Index, a benchmark that tests how well models handle agentic tasks. And you can run Qwen 3.8 27B locally on a single RTX 3090.
- · Alex 的文章 · Alex Getman
摘录
⚡️ Alibaba released Qwen3.8’s weights A model at Claude Opus 4.6 level, a flagship just six months ago, can now run on a single RTX 4090. Qwen3.8 27B uses around 17 GB of VRAM when quantized and should deliver ~50 tok/s.
- · Alex 的文章 · Alex Getman
摘录
Alibaba just teased Qwen3.8 with 2.4T parameters This is the new Qwen flagship, with open weights coming soon. Alibaba ranks it right below Claude Fable 5. Public benchmarks are still missing, so nobody really knows how close it actually is to Fable, or how it stacks up against GPT-5.6 Sol and Kimi K3. Pricing is the fun part. Kimi K3 costs $3/$15, so Qwen3.8 will probably land in the same range. The preview is already live in Qoder. It uses 10x fewer credits during the day and 50x fewer at night, so you can basically test it for free.
规格
Qwen3.8 · 已移出活跃追踪列表。尚未确认以此名称发布或更名。
常见问题
Qwen3.8 何时发布？
发布日期未确认