# Qwen3.8

Qwen3.8 · 已移出活跃追踪列表。尚未确认以此名称发布或更名。

- 开发商: Alibaba
- Id: qwen3.8
- 据报道: 2026-07-19T11:07:11.687Z
- 状态: 据报道
- 已移出活跃追踪列表。尚未确认以此名称发布或更名。
- 规范地址: https://alexgetman.com/zh/models/alibaba/qwen3-8/

## 常见问题

### Qwen3.8 何时发布？

发布日期未确认

## 时间线

- 2026-07-19T11:07:11.687Z · [Alex Getman](https://alexgetman.com/86/alibaba-just-teased-qwen3-8-with-2-4t-parameters/)
> 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.
- 2026-08-14T18:14:16.552Z · [Alex Getman](https://alexgetman.com/184/alibaba-open-sourced-qwen3-8/)
> ⚡️ 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.
- 2026-08-18T05:00:00.000Z · [Alex Getman](https://alexgetman.com/196/qwen-3-8-27b-beat-gpt-5-6-terra/)
> ⚡️ 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.
- 2026-08-26T11:23:48.753Z · [Alex Getman](https://alexgetman.com/356/qwen3-8-flash-next-drops-today/)
> 🚨 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.
- 2026-08-26T12:54:00.791Z · [Alex Getman](https://alexgetman.com/357/qwen3-8-flash-next-is-out/)
> 🚨 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.
- 2026-09-17T18:25:45.043Z · [Alex Getman](https://alexgetman.com/428/qwen3-8-omni-flash-is-out/)
> 🚨 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
- 2026-11-05T17:57:47.304Z · [GitHub · QwenLM/qwen-code · PR](https://github.com/QwenLM/qwen-code/pull/15184)
> - The Standard preset has the same qwen3.8 effort and GLM thinking gaps and will be aligned in a follow-up
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · 0xburden/modelship · @0xburden](https://github.com/0xburden/modelship/pull/8)
> - 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
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · Azure/azureml-assets · @hanbitmyths](https://github.com/Azure/azureml-assets/pull/5452)
> Add MAI, Qwen3.8 and Nemotron 3.5 Foundry Local coding models
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · BuQQzz/SHIELD2.0 · @BuQQzz](https://github.com/BuQQzz/SHIELD2.0/pull/56)
> Makes Qwen3.8 27B (IQ1_M) a working coding model on an 8 GB card (RTX 2070 SUPER) through SHIELD's own llama-server.
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · FlashML-org/FreeToken · @zihaomu](https://github.com/FlashML-org/FreeToken/pull/897)
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · NVIDIA/hyperflare-inference · @Raydius](https://github.com/NVIDIA/hyperflare-inference/pull/212)
> 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
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · SceneWorks/SceneWorks · @michaeltrefry](https://github.com/SceneWorks/SceneWorks/pull/3383)
> SW-7119: Qwen3.8 27B describer model and tier in core, worker, API and catalog
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · michsindlinger/specwright · @michsindlinger](https://github.com/michsindlinger/specwright/pull/125)
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · nearai/cvm-compose-files · @PierreLeGuen](https://github.com/nearai/cvm-compose-files/pull/488)
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vitessio/vitess · @jdoupe](https://github.com/vitessio/vitess/pull/21255)
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vjmichael/Lu-AI-GUI · @vjmichael](https://github.com/vjmichael/Lu-AI-GUI/pull/226)
> - 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.
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/2629)
> [Model][Qwen3.8] QSA main-KV decode context parallelism over rank pairs (DCP-2) kernels
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/2895)
> [Model][Qwen3.8] Qwen3.8-Flash model and registration hooks
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3232)
> [Model][Qwen3.8] QSA union-walk kernels (plan, dense, SparseCore gather, prefill)
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3233)
> [Model][Qwen3.8] QSA KV write kernels (fused, main, compressed)
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3234)
> [Model][Qwen3.8] Pooled QSA kernel over the unified KV pool
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3235)
> [Model][Qwen3.8] Hyper-connection kernels
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3236)
> [Model][Qwen3.8] PLE lookup and gate kernels
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3237)
> [Model][Qwen3.8] Unified KV pool kernels
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3238)
> [Model][Qwen3.8] GDN input preparation kernel
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3239)
> [Model][Qwen3.8] Torch bridge for the QSA kernels
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3240)
> [Model][Qwen3.8] QSA Torch-to-JAX ops
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3241)
> [Model][Qwen3.8] PLE Torch-to-JAX ops
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3242)
> [Model][Qwen3.8] In-place rows update op
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3243)
> [Model][Qwen3.8] Mamba tail fold: export prefill checkpoints from inside the step
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3244)
> [Model][Qwen3.8] Per-group Mamba index accessors
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3245)
> [Model][Qwen3.8] Model-owned GDN op and layer (no register_oot)
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3246)
> [Model][Qwen3.8] QSA attention layer
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3247)
> [Model][Qwen3.8] Sequence-parallel prefill backbone
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3248)
> [Model][Qwen3.8] PLE feature: tables, host lookup, storage, offload and pooled PLE
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3249)
> [Model][Qwen3.8] Config hook, geometry checks and pooled cache layout
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3250)
> [Model][Qwen3.8] PD transfer: identity, manifest and Raiden connector
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3251)
> [Model][Qwen3.8] Pooled KV cache materialization
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3252)
> [Model][Qwen3.8] GDN layer input preparation
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3253)
> [Model][Qwen3.8] Hyper-connection layers
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3254)
> [Model][Qwen3.8] Runner input adapter
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3255)
> [Model][Qwen3.8] KV offload pool regions
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3256)
> [Model][Qwen3.8] MTP draft model
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3267)
> [Model][Qwen3.8] qwen38_flash package entry: register() and READMEs
- 2026-11-13T19:07:08.264Z · [GitHub search · Alibaba · pulls · xadupre/devdash-agents · @xadupre](https://github.com/xadupre/devdash-agents/pull/19)
> - 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
- 2026-11-13T19:07:40.219Z · [GitHub search · Alibaba · commits · Azure/azureml-assets · @hanbitmyths](https://github.com/Azure/azureml-assets/commit/229f9a3bd592035925db8d0a49fd65f8b5c43998)
> Add MAI, Qwen3.8 and Nemotron 3.5 Foundry Local coding models (#5452)
- 2026-11-13T19:07:40.219Z · [GitHub search · Alibaba · commits · Hallo-ich-bin-PeppaPig/Singularity · @Hallo-ich-bin-PeppaPig](https://github.com/Hallo-ich-bin-PeppaPig/Singularity/commit/b33219c6ead9582f7e03467b2d6885718bbe16cc)
- 2026-11-13T19:07:40.219Z · [GitHub search · Alibaba · commits · QuantumKuba/Diffactly · @QuantumKuba](https://github.com/QuantumKuba/Diffactly/commit/77529540c23a8e7419a0147263f52ad0078ae3df)
- 2026-11-13T19:07:40.219Z · [GitHub search · Alibaba · commits · aabhay007/Personal-AI-Core · @aabhay007](https://github.com/aabhay007/Personal-AI-Core/commit/54e756490e80be233ca0314f9fed50baad560413)
- 2026-11-13T19:07:40.219Z · [GitHub search · Alibaba · commits · agentgateway/agentgateway · @github-actions[bot]](https://github.com/agentgateway/agentgateway/commit/b7f833ec371abdb5a58fe64c9a7409a51cd1d11c)
> json.providers.nebius.models["Qwen/Qwen3.8-27B"].name = "Qwen3.8 27B";
- 2026-11-13T19:07:40.219Z · [GitHub search · Alibaba · commits · fdietze/pi-infinite-context · @fdietze](https://github.com/fdietze/pi-infinite-context/commit/af207012a9c5522c75abcf01352ef15b166e983c)
- 2026-11-13T19:07:40.219Z · [GitHub search · Alibaba · commits · finereli/pyramid-mcp · @finereli](https://github.com/finereli/pyramid-mcp/commit/279f60c731b0ee6d5fe2b68bb2cfe24b97358c64)
- 2026-11-13T19:07:40.219Z · [GitHub search · Alibaba · commits · onprem-ai/deployment-catalog · @tomas-polach](https://github.com/onprem-ai/deployment-catalog/commit/13545a9aa19112ab1251fe153caa0cf872afddc1)

## 参考链接

- [Alex Getman](https://alexgetman.com/184/alibaba-open-sourced-qwen3-8/)
- [Alex Getman](https://alexgetman.com/196/qwen-3-8-27b-beat-gpt-5-6-terra/)
- [editorial](https://alexgetman.com/356/qwen3-8-flash-next-drops-today/)
- [editorial](https://alexgetman.com/357/qwen3-8-flash-next-is-out/)
- [editorial](https://alexgetman.com/428/qwen3-8-omni-flash-is-out/)
- [editorial](https://alexgetman.com/86/alibaba-just-teased-qwen3-8-with-2-4t-parameters/)
- [GitHub search · Alibaba · pulls · 0xburden/modelship · @0xburden](https://github.com/0xburden/modelship/pull/8)
- [GitHub search · Alibaba · commits · Azure/azureml-assets · @hanbitmyths](https://github.com/Azure/azureml-assets/commit/229f9a3bd592035925db8d0a49fd65f8b5c43998)
- [GitHub search · Alibaba · pulls · Azure/azureml-assets · @hanbitmyths](https://github.com/Azure/azureml-assets/pull/5452)
- [GitHub search · Alibaba · pulls · BuQQzz/SHIELD2.0 · @BuQQzz](https://github.com/BuQQzz/SHIELD2.0/pull/56)
- [GitHub search · Alibaba · pulls · FlashML-org/FreeToken · @zihaomu](https://github.com/FlashML-org/FreeToken/pull/897)
- [GitHub search · Alibaba · commits · Hallo-ich-bin-PeppaPig/Singularity · @Hallo-ich-bin-PeppaPig](https://github.com/Hallo-ich-bin-PeppaPig/Singularity/commit/b33219c6ead9582f7e03467b2d6885718bbe16cc)
- [GitHub search · Alibaba · pulls · NVIDIA/hyperflare-inference · @Raydius](https://github.com/NVIDIA/hyperflare-inference/pull/212)
- [GitHub search · Alibaba · commits · QuantumKuba/Diffactly · @QuantumKuba](https://github.com/QuantumKuba/Diffactly/commit/77529540c23a8e7419a0147263f52ad0078ae3df)
- [GitHub · QwenLM/qwen-code · PR](https://github.com/QwenLM/qwen-code/pull/15184)
- [GitHub search · Alibaba · pulls · SceneWorks/SceneWorks · @michaeltrefry](https://github.com/SceneWorks/SceneWorks/pull/3383)
- [GitHub search · Alibaba · commits · aabhay007/Personal-AI-Core · @aabhay007](https://github.com/aabhay007/Personal-AI-Core/commit/54e756490e80be233ca0314f9fed50baad560413)
- [GitHub search · Alibaba · commits · agentgateway/agentgateway · @github-actions[bot]](https://github.com/agentgateway/agentgateway/commit/b7f833ec371abdb5a58fe64c9a7409a51cd1d11c)
- [GitHub search · Alibaba · commits · fdietze/pi-infinite-context · @fdietze](https://github.com/fdietze/pi-infinite-context/commit/af207012a9c5522c75abcf01352ef15b166e983c)
- [GitHub search · Alibaba · commits · finereli/pyramid-mcp · @finereli](https://github.com/finereli/pyramid-mcp/commit/279f60c731b0ee6d5fe2b68bb2cfe24b97358c64)
- [GitHub search · Alibaba · pulls · michsindlinger/specwright · @michsindlinger](https://github.com/michsindlinger/specwright/pull/125)
- [GitHub search · Alibaba · pulls · nearai/cvm-compose-files · @PierreLeGuen](https://github.com/nearai/cvm-compose-files/pull/488)
- [GitHub search · Alibaba · commits · onprem-ai/deployment-catalog · @tomas-polach](https://github.com/onprem-ai/deployment-catalog/commit/13545a9aa19112ab1251fe153caa0cf872afddc1)
- [GitHub search · Alibaba · pulls · vitessio/vitess · @jdoupe](https://github.com/vitessio/vitess/pull/21255)
- [GitHub search · Alibaba · pulls · vjmichael/Lu-AI-GUI · @vjmichael](https://github.com/vjmichael/Lu-AI-GUI/pull/226)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/2629)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/2895)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3232)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3233)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3234)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3235)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3236)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3237)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3238)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3239)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3240)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3241)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3242)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3243)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3244)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3245)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3246)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3247)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3248)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3249)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3250)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3251)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3252)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3253)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3254)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3255)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3256)
- [GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG](https://github.com/vllm-project/vllm-torchtpu/pull/3267)
- [GitHub search · Alibaba · pulls · xadupre/devdash-agents · @xadupre](https://github.com/xadupre/devdash-agents/pull/19)
- [qwen-articles](https://qwen.ai/blog?id=8865c9dc-9833-4aa3-8c74-81c244bec7b0)

## 信号