Skip to content

Laguna S 2.1 and Laguna XS

Current status and review date

Decision snapshot

  • Product role: Laguna S is a retained rollback-era qualified recipe; Laguna XS is a rejected compatibility path. Neither label reports a current live route.
  • Selected or best-qualified configuration: Laguna S 2.1 NVFP4 with thinking disabled at 262,144 tokens; retained TP=1 and exclusive TP=2 evidence describe different measured lanes.
  • Measured hardware: one RTX PRO 6000 for the historical TP=1 lane and two equal RTX PRO 6000 cards over PCIe for the exclusive TP=2 refresh.
  • Evidence: Laguna S has functional, capacity, and quality evidence; Laguna XS has compatibility-only and historical-invalid evidence.
  • Decision: Laguna S was a dated rollback; the TP=2 refresh is no-promotion; Laguna XS is rejected under the tested recipes.
  • Important limitation: thinking-enabled Laguna S can exhaust its visible-answer budget, and some Laguna XS response bodies/full logs were not retained.
  • Review dates: retained evidence through 2026-08-01; dossier-format review 2026-08-31.

Open the retained recipe registry or jump to the decision, known limitations, or dated evidence.

Review narrative

2026-07-12 — Laguna XS engine A/B rejection

Laguna XS was attempted through vLLM and SGLang on one RTX PRO 6000. The vLLM FP8-KV lane exposed corrupted text and 0/20 tools; the non-FP8 fallback stalled during cache profiling. SGLang loaded but produced repetitive or off-topic output, missed the long-context and tool gates, and its documented trtllm_mha backend rejected sm_120. Some operator-observed healthy-run responses and full logs were lost when containers were recreated, so they are not reproducible quality claims.

Outcome: rejected for the tested configurations, not a judgment about the model on every Blackwell product.

2026-07-26 — Laguna S single-card qualification and promotion

Laguna S used vLLM 0.23.1rc1.dev1327+gf25953cc5, NVFP4 weights, FP8 KV, 262,144 context, and thinking disabled. Repeated protocol-v3 quality passed, including 32K, 128K, and 240K retrieval, tools, session recall, unified diff, and timeout triage. Thinking-enabled testing also showed an intermittent failure that consumed the complete 4,352-token completion allowance and returned no visible answer, so the qualified contract forced thinking off. The dated campaign then recorded a human-approved Heavy promotion with GPT-OSS Puzzle retained as rollback.

Outcome: the exact TP=1 disabled-thinking recipe qualified and was promoted in that dated campaign; this dossier does not claim it remains live.

2026-08-01 — Exclusive TP=2 refresh

The same exact Laguna S checkpoint ran on two equal PRO 6000 cards with pinned vLLM 0.25.1, NVFP4/FP8 KV, 262,144 context, one admitted request, no DFlash, and thinking disabled. Smoke, JSON, 30K retrieval, tools, extended tools, intelligence 6/6, session 3/3, and tools 3/3 passed. At 32K it measured 1.97-second TTFT, 15,134 effective prefill tok/s, and 70.9 tok/s decode. At a 231,457-token prompt it measured 31.85-second TTFT, 7,252 effective prefill tok/s, and 66.0 tok/s decode.

Outcome: qualified as a TP=2 no-promotion comparison lane; no production alias or router profile changed.

Immutable identity

Laguna S 2.1

  • Model: poolside/Laguna-S-2.1-NVFP4 revision 07614121b31898586430f189d27a25a0be310843.
  • TP=1 served name: laguna-s-2.1-nvfp4.
  • TP=2 served name: laguna-s-2.1-nvfp4-tp2.
  • TP=1 runtime: vLLM 0.23.1rc1.dev1327+gf25953cc5, image reference vllm/vllm-openai:nightly-f25953cc59f9b4ba9b04b16228d2b86dcfbcbdb1.
  • TP=1 registry digest: Not retained; the engine-revision tag is not a registry digest.
  • TP=2 runtime: vLLM 0.25.1, image sha256:e4f88a835143cd22aee2397a26ec6bb80b3a4a6fe0c882bcbc63822904766089.

Laguna XS 2.1

  • Model: poolside/Laguna-XS-2.1-NVFP4 revision 07133fb3df1cc3111478e24ee71a823a598c8c2f in the retained July 12 evaluation.
  • Qualification: no qualified served recipe resulted.
  • Retained runtimes: vLLM image sha256:907377dddef392f6b679d9c071e1c33c3935b4dc993b61d0352e391a5319ff3e and SGLang image sha256:0261d11a9bf5ef8041d68d55f0bac3f3375330e21353245e23ef461ffbe57da5.

Tested hardware and topology

Laguna S TP=1

  • Measured: one RTX PRO 6000 Blackwell 96 GB on an isolated endpoint.
  • Execution mode: single-card, five or fewer admitted requests depending on the workload; the promotion contract served thinking disabled.

Laguna S TP=2

  • Measured: two equal RTX PRO 6000 Blackwell Max-Q cards.
  • Execution mode: exclusive TP=2 over PCIe without NVLink, concurrency one.
  • Comparability boundary: runtime, topology, admission, and workload differ from the TP=1 lane, so this is not a topology-only speed A/B.

Laguna XS compatibility probes

  • Measured: one RTX PRO 6000, isolated candidate endpoints.
  • Boundary: the failures apply to the exact tested vLLM/SGLang recipes.

Engine, quantization, KV, context, and concurrency recipe

Laguna S TP=1 rollback-era recipe

  • Engine and image: vLLM 0.23.1rc1.dev1327+gf25953cc5, commit-tagged nightly image above.
  • Weights and KV: compressed-tensors NVFP4 and FP8 KV.
  • Contract: 262,144 tokens; request-level chat_template_kwargs.enable_thinking=false.
  • Recipe: single-card registry entry.

Laguna S TP=2 refresh

  • Engine and image: pinned vLLM 0.25.1 image digest above.
  • Weights and KV: NVFP4 and FP8 KV.
  • Contract: TP=2, 262,144 tokens, one admitted request, no DFlash, thinking disabled.
  • Recipe: TP=2 campaign registry.

Laguna XS rejected probes

  • vLLM: FP8-KV and non-FP8 fallback probes only.
  • SGLang: Laguna-specific CUDA 13 image with default and explicit template probes.
  • Recipe: retained experimental Compose.

Evidence by measurement class

Laguna S TP=1

  • Status: functional, capacity, and quality.
  • Measured: retrieval passed at 32K, 128K, and 240K targets; repeated tools/session/intelligence checks passed. Short-output capacity completed 10/10 at c1 and 40/40 at c8, with 75.46 and 83.24 aggregate tok/s.
  • Evidence: qualification finding and its linked raw artifacts.

Laguna S TP=2

  • Status: functional, capacity, and quality; no-promotion.
  • Measured: functional and repeated gates passed; 32K measured 1.97-second TTFT, 15,134 prefill tok/s, and 70.9 tok/s decode; 231,457 prompt tokens measured 31.85-second TTFT, 7,252 prefill tok/s, and 66.0 tok/s decode.
  • Evidence: campaign finding, 32K artifact, and 240K artifact.

Laguna XS

  • Status: compatibility-only and historical-invalid; tested recipes rejected.
  • Measured: the retained negative evidence includes corrupted text, 0/20 tools, startup stalls, empty long-context output, and backend rejection.
  • Limit: full response bodies/logs for some operator-observed attempts were not retained.
  • Evidence: engine A/B finding and machine-readable failure.

Decision and promotion state

Retained or selected

  • Laguna S TP=1: dated rollback-era qualified recipe. Do not interpret this as a current live assignment.
  • Laguna S TP=2: no-promotion; retained as a qualified comparison lane.

Rejected

  • Laguna XS: rejected under the tested vLLM and SGLang configurations; reconsideration requires fresh immutable identity and complete evidence.

Failures and gotchas

Reasoning and evidence limits

  • Laguna S thinking: the qualified contract disables thinking because the enabled path intermittently exhausted the completion budget without a visible answer.
  • Laguna XS retention gap: some response bodies and complete logs were not retained, so those attempts remain historical-invalid rather than quality results.

Runtime and topology limits

  • Laguna XS FP8 KV: produced corrupted output and 0/20 tools.
  • Laguna XS fallback: non-FP8 vLLM stalled; SGLang output remained repetitive/off-topic and its documented trtllm_mha path rejected sm_120.
  • Cross-lane comparison: TP=1 and TP=2 results use different runtimes, admission, and workloads.

Dated run history