{"slug":"ds4-flash-2x6000-jj-dspark-k5","title":"DeepSeek V4 Flash-0731 on 2x RTX PRO 6000 Blackwell \\x{2014} TP2 DSpark K5","summary":"Community image TP2 DSpark K5 serving recipe: B12X sparse MoE, FP8 KV, InstantTensor loading, 1M context. Saved numbers from the KK comparison (Max-Q +6000).","license":"CC-BY-4.0","tier":"community","beta":false,"revision":{"no":2,"commit_sha":"aa77cf295a865cd2380aa3150da706889728140b","content_hash":"655e2c0c651bd72eb6a493c12824a232b005a10c","created_at":"2026-10-20 21:54:52.587962-06","notes":"Compose file pulls the community r9 image (MODE dspark, DSPARK_TOKENS 5, MAX_NUM_SEQS 8)."},"engine_source":{"type":"docker_image","git_url":"https://github.com/voipmonitor/vllm/tree/integration/jovian-reviewed-sources-20260911","base_engine":"vllm","pip_package":null,"base_version":"jovian-judgement r9 (vllm f66599d / b12x 15b6813)","image_digest":"voipmonitor/vllm:jovian-judgement-vllmf66599d-b12x15b6813-fi803c466-cu133-torch213-20260907-r9","install_note":null,"pinned_commit":null},"credit":[{"name":"local-inference-lab","url":"https://github.com/local-inference-lab/rtx6kpro"}],"author":{"name":"localai-recipes","isSite":true},"uploaded_by":"localai-recipes","config":{"launch":["docker","compose","-f","docker-compose-ds4-dspark-jovian-judgement-r9.yml","up","-d"]},"models":[{"slug":"deepseek-v4-flash-0731","name":"DeepSeek-V4-Flash-0731","hf_repo":null,"quant":"fp8"}],"hardware":[{"slug":"nvidia-rtx-pro-6000-blackwell-max-qx2","vendor":"nvidia","card":"RTX PRO 6000 Blackwell Max-Q","card_family":"rtx-pro-6000-blackwell","vram_gb":96,"device_count":2}],"generation":"blackwell","tags":["dspark","jovian-judgement"],"tasks":["chat","code"],"aa":{"aa_index":34,"aa_provenance":"measured","aa_source_url":"https://artificialanalysis.ai/models/deepseek-v4-flash","aa_index_version":"Index v4.3.2","aa_methodology_version":"v4.3.2","aa_published_fetched_at":"2026-10-07 18:00:00-06","aa_family_slug":"deepseek-v4-flash-0731","aa_resolved_slug":"deepseek-v4-flash-0731"},"performance":[{"claimed_by":"submitter","metrics":"{\"points\": [{\"notes\": \"Saved JJ (KK comparison), 5-window median, ctx-0 decode\", \"concurrency\": 1, \"context_len\": 1, \"decode_tok_s\": 214.900000000000005684341886080801486968994140625}, {\"notes\": \"Saved JJ, aggregate C8\", \"concurrency\": 8, \"context_len\": 1, \"decode_tok_s\": 611.8999999999999772626324556767940521240234375}, {\"notes\": \"Saved JJ, uncached nominal-32K prefill\", \"concurrency\": 1, \"context_len\": 32768, \"prefill_tok_s\": 11167}]}"}],"perf_points":[{"id":1735,"context_len":1,"prefill_tok_s":null,"decode_tok_s":214.900000000000005684341886080801486968994140625,"ttft_ms":null,"concurrency":1,"claimed_by":"submitter","notes":"Saved JJ (KK comparison), 5-window median, ctx-0 decode"},{"id":1736,"context_len":1,"prefill_tok_s":null,"decode_tok_s":611.8999999999999772626324556767940521240234375,"ttft_ms":null,"concurrency":8,"claimed_by":"submitter","notes":"Saved JJ, aggregate C8"},{"id":1737,"context_len":32768,"prefill_tok_s":11167,"decode_tok_s":null,"ttft_ms":null,"concurrency":1,"claimed_by":"submitter","notes":"Saved JJ, uncached nominal-32K prefill"}],"agent":{"prompt":"Fetch https://localairecipes.com/api/v1/recipes/ds4-flash-2x6000-jj-dspark-k5.json and apply the inference recipe it describes (engine flags, model, quant, context) to my setup. Ask before overwriting any existing config.","curl":"curl -fsSL https://localairecipes.com/api/v1/recipes/ds4-flash-2x6000-jj-dspark-k5.json"},"requirements":{"requires":["Docker (or a compatible OCI container runtime)","NVIDIA Container Toolkit (nvidia-container-toolkit) - exposes host GPUs to the container"],"driver_floor":{"cuda":"13.3","linux_min":">=580","text":"NVIDIA driver >= 580 (image CUDA 13.3 - CUDA 13.x applications run on drivers >= 580 under CUDA minor-version compatibility; the CUDA 13.3 toolkit shipped driver >= 610.43.02)","source":"NVIDIA CUDA Toolkit Release Notes - driver compatibility tables (https://docs.nvidia.com/cuda/cuda-toolkit-release-notes/, fetched 2026-10-03)"},"placeholders":["Hardware-bound values in this recipe (device counts, bind-mount paths, utilization, ports) describe the author's machine, not yours - map them to your own rig before launching. Keep the engine flags, model, quant, and parallelism strategy; adjust paths and ports to the hardware you actually have.","GPU identity: list your GPUs with `nvidia-smi -L` and substitute your own GPU UUIDs/indices wherever the recipe pins the author's.","Model directory: bind-mount paths are author-side - point them at your own mount where the model files live.","Host port: pick a free port on your machine and publish the container there; keep the in-container port unchanged."]},"artifacts":[],"links":{"self":"/api/v1/recipes/ds4-flash-2x6000-jj-dspark-k5.json","alternate":"/r/ds4-flash-2x6000-jj-dspark-k5","describedby":"/openapi.json"},"updated_at":"2026-10-20 21:54:52.587962-06","created_at":"2026-10-20 21:49:12.162484-06","speeds":{"decode_tok_s":214.900000000000005684341886080801486968994140625,"prefill_tok_s":11167,"leaderboard":"/api/v1/leaderboard.json"}}