SHANGHAI AI LABORATORY / INTERNLM · FIELD VERIFIED
Sage S2.1 Flash
Primary sources last checked on 19 Nov 2026. Verification freshness is tracked separately from factual change, and unknown values remain explicit rather than inferred.
NotableOfficial publisher original checkpoint is publicly accessible at an immutable 40-character revision, with concrete visible weights and a source-specified license. Independent OMW runtime or benchmark performance has not been established; Notable only.Curation method →
Open Model Weights records Sage S2.1 Flash as a 321B publisher-reported total parameters-parameter scientific multimodal input, text and tool-use output open-weight model in the Sage S2.1 family, with Shanghai AI Laboratory / InternLM listed as the developer. The checked source repository provides 4 recognized weight artifacts, format evidence for Safetensors and a context window that is not declared in the checked standard sources. Values that are not evidenced in the checked sources remain explicitly undisclosed or unknown.
The repository declares MIT. Commercial use allowed under mit license conditions. Open Model Weights reports the checked license evidence and classification; this is not legal advice.
No base model is declared in the checked standard lineage metadata, so Open Model Weights does not infer one. Training recipe status: publisher-described-not-reproduced; data disclosure status: Not verified in the checked first-party material.
Shanghai AI Laboratory's 321B Sage S2.1 Flash is a source-pinned open-weight scientific multimodal and tool-use model. The official repository exposes safetensors shards and claims MIT in its License section, but does not supply standardized Hugging Face YAML license metadata; that evidence distinction is preserved. The model profile links an exact original-repository revision and visible files; hardware and benchmark performance remain untested by Open Model Weights.
Official publisher sourceinternlm/Sage-S2.1-FlashSource ↗
Original weight evidence4 explicitly recorded source-visible Safetensors path(s); limited subset of a larger numbered shard familySource ↗
License provenancemit — explicit official model-card section; not Hugging Face YAML metadataSource ↗
Parameter and context claim321B publisher-reported total parameters; no source-confirmed contextSource ↗
Benchmark statusPublisher or external leaderboard claims only; no independent OMW inference or benchmarkSource ↗
WHY IT MATTERS
Scientific input pathways cover molecular graphs, crystal structures and protein sequences/3D data through Sage SciPerception projectors; tool use supports multi-step research workflows.
The official file browser visibly contains model-00001-of-00120.safetensors and other shard filenames. The recorded subset is not represented as a full independent audit of all 120 shards.
The developer's model-card License section says code and model weights are MIT licensed even though the hub summary warns that YAML metadata is missing; the model profile cites that primary-source card, not inferred metadata.
MODEL-SELECTION VIEW
Strong fit when…
Evaluating scientific multimodal reasoning grounded in physical and biological inputs.
Building research agent pilots with traceable tool-use histories.
Source-specific comparisons of large scientific MoEs without inventing independent rank.
Important considerations
The 321B figure is a publisher/model-card total, not an independently measured tensor count.
Repository requires compatible projector and runtime components; a weight shard alone is not a complete turnkey scientific pipeline.
Benchmark rankings on the model card are reported by the publisher and not separately reproduced by OMW.
Model purpose and architecture
Shanghai AI Laboratory's 321B Sage S2.1 Flash is a source-pinned open-weight scientific multimodal and tool-use model. The official repository exposes safetensors shards and claims MIT in its License section, but does not supply standardized Hugging Face YAML license metadata; that evidence distinction is preserved.
Scientific input pathways cover molecular graphs, crystal structures and protein sequences/3D data through Sage SciPerception projectors; tool use supports multi-step research workflows.
Pinned public weight artifacts
The official file browser visibly contains model-00001-of-00120.safetensors and other shard filenames. The recorded subset is not represented as a full independent audit of all 120 shards.
The checked immutable source revision is 76f4513db7464485605a002a06a6190de0460574. Only the explicitly listed first-party shard paths were captured in this manual review; the full numbered repository shard family was not independently enumerated. A file name is not a measured runtime result.
License and provenance
The developer's model-card License section says code and model weights are MIT licensed even though the hub summary warns that YAML metadata is missing; the model profile cites that primary-source card, not inferred metadata.
License evidence: the official publisher card, with missing standardized YAML metadata under mit. The checkpoint license does not automatically cover third-party datasets, application code or generated outputs.
Performance evidence and limitations
These are publisher or independently hosted third-party benchmark claims, not tests Open Model Weights has executed. A fair comparison requires an exact checkpoint, matched prompts, harness, hardware and repeatable samples.
The 321B figure is a publisher/model-card total, not an independently measured tensor count.
Deployment and practical evaluation
Appropriate evaluations include: Evaluating scientific multimodal reasoning grounded in physical and biological inputs. Building research agent pilots with traceable tool-use histories. Source-specific comparisons of large scientific MoEs without inventing independent rank. Source code and quantizations from separate repositories must be checked independently.
Repository requires compatible projector and runtime components; a weight shard alone is not a complete turnkey scientific pipeline. Benchmark rankings on the model card are reported by the publisher and not separately reproduced by OMW.
IMPLEMENTATION & EVALUATION
How to evaluate Sage S2.1 Flash before deployment
Primary-source pinned checkpoint evaluation steps; no independent OMW model inference performed.
STEP 01
Pin official checkpoint and artifact paths
Publisher repositories can change after publication.
REPRODUCIBLE ACCEPTANCE TEST
Record the source 40-character revision and verified visible filenames, along with processor/config revision.
WHAT CAN GO WRONG
Mutable tags, missing shards and silent upstream changes.
STEP 02
Review license and code dependencies
Open model weights do not imply unrestricted use of code or all data.
REPRODUCIBLE ACCEPTANCE TEST
Inspect the exact publisher license source, custom modeling code, dependencies and intended use policy.
WHAT CAN GO WRONG
Misclassified commercial license or remote-code execution.
STEP 03
Measure hardware, context and workload
Advertised parameters, context and token rates are not observed system requirements.
REPRODUCIBLE ACCEPTANCE TEST
Publish repeatable measured memory, batch, hardware, latency and context settings.
WHAT CAN GO WRONG
Out-of-memory, unsupported attention operators and overclaimed throughput.
STEP 04
Separate external claims from OMW results
Publisher comparisons have different harnesses.
REPRODUCIBLE ACCEPTANCE TEST
Use held-out tasks and matched baselines before independent ranking.
WHAT CAN GO WRONG
False benchmarking attribution or ungrounded Leading status.
Are the original Sage S2.1 Flash model weights publicly available?
Yes. The official source repository internlm/Sage-S2.1-Flash exposes actual Safetensors checkpoint artifacts at the pinned revision 76f4513db7464485605a002a06a6190de0460574. This review lists source-visible example weight paths, not the entire multi-shard manifest.
Is Sage S2.1 Flash independently benchmarked by Open Model Weights?
No. Published scores and throughput descriptions are attributed to their source; OMW has not run model inference or a controlled benchmark.
What license applies to this exact checkpoint?
mit. Evidence is the official model-card License section; follow exact terms and separately inspect dependencies.
Is the model suitable for local deployment?
Original weights are downloadable, but actual fit, operators and performance need hardware-specific testing. Repository requires compatible projector and runtime components; a weight shard alone is not a complete turnkey scientific pipeline.
Does Notable mean the model leads an independent benchmark?
No. Notable records a distinct useful, source-verifiable original checkpoint. Leading requires separate matched independent evidence.
03 · PRACTICAL TECHNICAL PROFILE
Deployment, licensing and reproducible evidence.
Inspect weight-memory estimates, model lineage and runtime evidence separately from promotional or untested performance claims.
DEPLOYMENT PROFILE
Weight memory, without pretending it is a deployment guarantee.
These bars visualize weight-only decimal-GB estimates. They exclude KV cache, activations, runtime overhead, optimizer state and sharding.
WEIGHT-ONLY MEMORY
BF16 / FP16~642.0 GB
FP8 / INT8~321.0 GB
INT4~160.5 GB
FIELD EVIDENCE MAP
WeightsVerified
LicenseVerified
Commercial useAllowed
ContextNot declared
FormatsVerified
LineageNot declared
Training recipePublisher described not reproduced
Training dataNot verified in the checked first party material
RuntimeSource-derived · not runtime-tested
LICENSE & USE
MIT
Commercial use allowed under mit license conditions
Declared licenseMIT
ClassificationAllowed
VerificationPublisher card license checked
This section summarizes the checked repository evidence. It is not legal advice. Inspect license evidence ↗
LINEAGE & PROVENANCE
What the checked sources disclose — and what they do not.
Open Model Weights does not infer a hidden base model or training corpus when the checked sources do not declare one.
LINEAGE
BASE / PARENTNot declared in model-card metadataUnknown
not inferred
CURRENT RECORDSage S2.1 FlashSage S2.1
TRAINING DISCLOSURE
Training recipePublisher described not reproduced
Training dataNot verified in the checked first party material
FamilySage S2.1
RUNTIME SIGNALS
Compatibility claims stay attached to their evidence level.
Declared and model-card-mentioned runtimes are discovery signals. They are not treated as Open Model Weights runtime tests unless explicitly measured.
MentionedvLLM
OMW runtime testNot performed
EVIDENCE TRAIL
Sources and observed history.
This is where the article remains auditable: source links on one side, the registry's observed change history on the other.
Three evidence-linked visual summaries: parameter scale, illustrative precision memory and verifiable provenance. No performance ranking is implied.
01 / PARAMETER SCALE
Total versus active parameters
Total parameters321B
Active parameter count not established.
Repository-derived or publisher-described parameter count
Published total parameters are not a model-quality or performance score.
02 / MEMORY ESTIMATE
Weight-only precision footprint
BF16 / FP16642 GB
FP8 / INT8321 GB
INT4160.5 GB
Hypothetical weight-only decimal GB, not measured peak VRAM or guaranteed support for each format. KV cache, activations, runtime and sharding are excluded.
Repository SHA 76f4513db746…; a source check is not an inference benchmark or legal clearance. See the per-model evidence JSON for the exact source record.
A verification check does not pretend the model changed. Open Model Weights updates the verification date after a real primary-source check, while factual-update dates move only when retained source evidence changes.
Declared, mentioned and externally evidenced runtimes
Declared libraryNot declared
Declared/taggedNone
Mentioned in model cardvLLM
Artifact signalsNone
External primary runtime evidenceNone
Tested by Open Model WeightsNo — source-derived support only
Open Model Weights separates model-repository evidence from external runtime-release evidence. An external support statement does not change the official model repository's weight formats and does not imply an Open Model Weights inference benchmark.
training recipesource-described-not-independently-reproduced
data disclosureNot verified in the checked first-party material
runtime supportsource-derived-not-runtime-tested
MODELS FOR AI AGENTS · ROLE EVIDENCE
Where this model fits in an agent system
Specialized model. A supporting component. This checkpoint does not, by category alone, serve as the planning and tool-execution controller.
The curated notable tier is not an agent-readiness rating. Model-specific tool-call success, long-horizon task accuracy and compatible runtime configurations are not independently established here.
Source-linked license: MIT (allowed commercial-use classification; see original terms).
Context: Unknown or not declared.
BF16 weights: 642 estimated decimal GB, not runtime VRAM.