Qwen3.8
Alibaba
타임라인
- · 근거 링크 · GitHub search · Alibaba · commits · onprem-ai/deployment-catalog · @tomas-polach
- · 근거 링크 · GitHub search · Alibaba · commits · finereli/pyramid-mcp · @finereli
- · 근거 링크 · GitHub search · Alibaba · commits · fdietze/pi-infinite-context · @fdietze
- · Alex의 게시물 · 제품군 · GitHub search · Alibaba · commits · agentgateway/agentgateway · @github-actions[bot]
발췌
json.providers.nebius.models["Qwen/Qwen3.8-27B"].name = "Qwen3.8 27B";
- · 근거 링크 · GitHub search · Alibaba · commits · aabhay007/Personal-AI-Core · @aabhay007
- · 근거 링크 · GitHub search · Alibaba · commits · QuantumKuba/Diffactly · @QuantumKuba
- · 근거 링크 · GitHub search · Alibaba · commits · Hallo-ich-bin-PeppaPig/Singularity · @Hallo-ich-bin-PeppaPig
- · 근거 링크 · GitHub search · Alibaba · commits · Azure/azureml-assets · @hanbitmyths
발췌
Add MAI, Qwen3.8 and Nemotron 3.5 Foundry Local coding models (#5452)
- · Alex의 게시물 · 제품군 · GitHub search · Alibaba · pulls · xadupre/devdash-agents · @xadupre
발췌
- 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
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] qwen38_flash package entry: register() and READMEs
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] MTP draft model
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] KV offload pool regions
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] Runner input adapter
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] Hyper-connection layers
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] GDN layer input preparation
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] Pooled KV cache materialization
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] PD transfer: identity, manifest and Raiden connector
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] Config hook, geometry checks and pooled cache layout
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] PLE feature: tables, host lookup, storage, offload and pooled PLE
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] Sequence-parallel prefill backbone
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] QSA attention layer
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] Model-owned GDN op and layer (no register_oot)
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] Per-group Mamba index accessors
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] Mamba tail fold: export prefill checkpoints from inside the step
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] In-place rows update op
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] PLE Torch-to-JAX ops
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] QSA Torch-to-JAX ops
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] Torch bridge for the QSA kernels
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] GDN input preparation kernel
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] Unified KV pool kernels
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] PLE lookup and gate kernels
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] Hyper-connection kernels
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] Pooled QSA kernel over the unified KV pool
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] QSA KV write kernels (fused, main, compressed)
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] QSA union-walk kernels (plan, dense, SparseCore gather, prefill)
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] Qwen3.8-Flash model and registration hooks
- · 근거 링크 · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
발췌
[Model][Qwen3.8] QSA main-KV decode context parallelism over rank pairs (DCP-2) kernels
- · 근거 링크 · GitHub search · Alibaba · pulls · vjmichael/Lu-AI-GUI · @vjmichael
발췌
- 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.
- · 근거 링크 · GitHub search · Alibaba · pulls · vitessio/vitess · @jdoupe
- · 근거 링크 · GitHub search · Alibaba · pulls · nearai/cvm-compose-files · @PierreLeGuen
- · 근거 링크 · GitHub search · Alibaba · pulls · michsindlinger/specwright · @michsindlinger
- · Alex의 게시물 · 제품군 · GitHub search · Alibaba · pulls · SceneWorks/SceneWorks · @michaeltrefry
발췌
SW-7119: Qwen3.8 27B describer model and tier in core, worker, API and catalog
- · Alex의 게시물 · 제품군 · GitHub search · Alibaba · pulls · NVIDIA/hyperflare-inference · @Raydius
발췌
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
- · 근거 링크 · GitHub search · Alibaba · pulls · FlashML-org/FreeToken · @zihaomu
- · Alex의 게시물 · 제품군 · GitHub search · Alibaba · pulls · BuQQzz/SHIELD2.0 · @BuQQzz
발췌
Makes Qwen3.8 27B (IQ1_M) a working coding model on an 8 GB card (RTX 2070 SUPER) through SHIELD's own llama-server.
- · 근거 링크 · GitHub search · Alibaba · pulls · Azure/azureml-assets · @hanbitmyths
발췌
Add MAI, Qwen3.8 and Nemotron 3.5 Foundry Local coding models
- · 근거 링크 · GitHub search · Alibaba · pulls · 0xburden/modelship · @0xburden
발췌
- 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
- · 근거 링크 · GitHub · QwenLM/qwen-code · PR
발췌
- The Standard preset has the same qwen3.8 effort and GLM thinking gaps and will be aligned in a follow-up
- · Alex의 게시물 · 제품군 · Alex Getman
발췌
🚨 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
- · Alex의 게시물 · 제품군 · Alex Getman
발췌
🚨 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.
- · Alex의 게시물 · 제품군 · Alex Getman
발췌
🚨 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.
- · Alex의 게시물 · Alex Getman
발췌
⚡️ 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.
- · Alex의 게시물 · Alex Getman
발췌
⚡️ 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.
- · Alex의 게시물 · Alex Getman
발췌
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.
사양
레이더 · Qwen3.8 · 출시일 미확인
자주 묻는 질문
Qwen3.8의 출시일은 언제인가요?
출시일 미확인