Model overview
Complete copy-paste recipe — pull the container, pull the model (fresh), pull the DFlash drafter (fresh), then serve with the vetted DGX Spark recipe (DFlash n=10, drafter `flash_attn`, body `triton_attn`). The recipe notes below cover tag-specific drafter backends and the production GPU-util / context-tier profile; see also Container Image Details. **Recipe notes — the drafter backend is tied to the image tag:** - **On `:latest` / `:2026-07-01-v0.24.0` (current; same behavior as the `:2026-06-18-v0.23.0-dflashfix` fix lineage):** the drafter **must** use `attention_backend: flash_attn` (enabled for multimodal Gemma targets by PR #41703's `use_mm_prefix=False`). `flex_attention` **crashes at the first request** on this image (non-contiguous KV view in the new KV-sharing path). `--enable-prefix-caching` is safe and recommended — soak-validated with DFlash (see the fixes section below). - **On the rollback tag `:2026-06-04-pr44389` (pre-fix):** the reverse — only `flex_attention` loads (`flash_attn` fails with `partial multimodal token full attention not supported`), the drafter silently runs without sliding-window support (~0% acceptance beyond 2k-token contexts), and `--enable-prefix-caching` triggers a slow acceptance collapse (see below). Not recommended. - Either way the body stays on `triton_attn` (Gemma's heterogeneous head dims), and do **not** set `--kv-cache-dtype fp8` — DFlash's non-causal drafter requires BF16 KV. `--reasoning-parser gemma4` cleanly splits `enable_thinking` output into `reasoning_content`. `num_speculative_tokens=10` is the value our production fleet runs; the drafter's trained block size is 16 and z-lab ships 15 — an n=10-vs-15 re-bench on the fixed drafter is pending (the earlier "10 beats 15" result was measured on the pre-fix drafter and shouldn't be treated as settled). The model's `config.json` already ships the vision ignore-list fix (`re:.*embed_vision.*`, `re:.*vision_tower.*`) so vLLM keeps the vision tower BF16. This default profile suits agentic gateways — one large full-context chat plus many short-lived subagents under the `--max-num-seqs 64` cap. For short-context throughput benchmarking, use `--max-model-len 32768 --max
Key features
特点摘要待依据来源补齐，不凭模型名称推断能力。
Limitations & considerations
Sources are insufficient to establish specific limitations. Verify task performance, latency, data policies, licensing and version lifecycle.
Source prices & cost analysis
Prices vary by channel, region and configuration. Channel quotes are listed independently and are not treated as a single official provider price.
| Source / channel | Input | Output | Billing unit | Collected at |
|---|---|---|---|---|
| No verified prices yet. Check the source channel. | ||||
Estimate monthly base token costs
Estimates exclude cache discounts, tiered pricing, multimodal surcharges, web search, taxes and additional reasoning output.
Public benchmarks & run configuration
暂无已采集的可核验公开评分。模型未上榜不代表能力较低。
View full indexed rankings →Frequently asked questions
Do open weights permit free commercial use?
Not necessarily. Check the version license, commercial conditions and usage policies. Self-hosting incurs compute and maintenance costs.
Are these real-time billing prices?
These are timestamped source snapshots, not live billing prices. Check the channel for current versions, regions, discounts and billing methods.
Sources & traceability
许可证：gemma。模型介绍与第三方评测均以原来源的具体版本和方法为准。