AI 模型库

Qwen3.8-Flash-Next is 360 GB in BF16 and needs two data-center GPUs in FP8. This build serves it on one 96 GB Blackwell : 88.8 GiB of VRAM, the 51B n-gram table in host RAM, and no patched runtime. Two flags are load-bearing on a single GPU. `--distributed-executor-backend mp`: the default single-GPU executor never starts the n-gram offload worker, and the first forward waits on it forever — the server looks healthy and hangs. `VLLM_PLE_OFFLOAD_READY_TIMEOUT=1800`: the worker loads a 95 GB table before serving and the 600 s default can expire first. Host needs about 100 GB of free RAM. The reasoning parser and tool-call parser in this block are validated on this exact checkpoint: thinking lands in `reasoning` with no markup in `content`, and tool calls arrive as structured `tool_calls` with valid JSON arguments and `finish_reason: tool_calls`. If you have seen other NVFP4 cards warn that the automatic MoE backend picks `FLASHINFER_TRTLLM` and corrupts output: on this image on an RTX PRO 6000 the automatic choice is `FLASHINFER_CUTLASS`, printed at boot, and every number on this card went through it. Measured on the mixed sibling (same expert kernel), `--moe-backend marlin` decodes 10% faster single-stream (93.2 vs 84.6 tok/s) and 1–3% slower at concurrency 32, with TTFT up 12%; tool-calling under Marlin 79.5 (n=1). Marlin does not combine with MTP on this image: the BF16 MTP draft inherits the backend and refuses it, and the draft-scoped override in `--speculative-config` is ignored by Model Runner V2 until vllm-project/vllm#54788 lands. Choose MTP for interactive use, Marlin only without speculation. On vLLM nightly builds from early September the SM120 selector picks Marlin by default anyway (vllm#54788); the flag matters on this image. Two more flags are in the block since 2026-09-06: without them the image's defaults do not boot on a 96 GB card. A field report found it first and we reproduced both failures on this box the same day. Without MTP the default `--max-num-seqs` (1024) exceeds the Mamba state blocks left at 0.92 utilization (598 here) and CUDA graph capture aborts with `Engine core initialization failed`; with MTP at the native 262,1

Visual QA

Input Text · Image→ Output Text

Price not verifiedCheck multimodal and channel pricing separately

View details ↗

CompareSource / docs ↗

Find your ideal AI model.

Explore AI models 10

K2-Horizon-MoVA-36B-A4B-NVFP4

LFM2.5-24B-A2B-NVFP4A16

Muse-Glimmer-30B-mixed-NVFP4

Nanbeige4.3-33B-FP8

Nanbeige4.3-33B-NVFP4

Nemotron-3.5-Lightning-30B-A3B-mixed-INT4-INT8

Ornith-1.5-35B-A3B-agentic-NVFP4-FP8

Qwen3.8-27B-mixed-NVFP4-FP8

Qwen3.8-Flash-Next-mixed-NVFP4-FP8

Qwen3.8-Flash-Next-NVFP4

No matching models

Great models. Clear costs.

Understand before you choose

Follow trends. Find tools. Explore models.