Nemotron 3 Nano 30B A3B vs Step 3.7 Flash

Nemotron 3 Nano 30B A3B 80

The model card earns 20, with all seven items on the card and its subcards. Training data earns 20, datasets listed with sources and token counts, both cut-offs, the filtering, and much of the data released, though some synthetic sets don't give a size. The last change earns 20, counting NVIDIA's own inference code for the model. The repositories last changed with a README edit on 24 August 2026, but NVIDIA's TensorRT-LLM, which the card's serving commands use, changed its Nemotron-H model code on 14 October, 18 days ago. Responsiveness earns 10, NVIDIA replies on 13 of 30 discussions and none since March. Version policy earns 5 for a first release with no dated notes. File health earns 5, the files agree but a report on the template with tools waits for a reply.

Step 3.7 Flash 39

Model card 14, Intended use, the architecture, the parameters, the context and the language (in the metadata) are on the card. Training data 0, Nothing on the card about the training data. Last change 10, the last commit, on 3 June 2026, fixed the modelling code, 160 days before this check and within 180. Responsiveness 0, none of the 11 most recent discussions that aren't pull requests (29 May to 24 June 2026) has a reply from a stepfun-ai member, counted on 10 November by author flags and dates with a script, text not read. Version policy 5, Step-3.5-Flash stays in its own repositories and 3.7 came out in new ones, with no dated note on the older cards that we read. File health 10, Config, template and card agree on the architecture and the context, and June commits fixed the modelling code. 14 + 0 + 10 + 0 + 5 + 10 = 39.

Nemotron 3 Nano 30B A3B 90

Memory earns 25 because the smallest 4-bit build is Ollama's NVFP4 MLX build at 18 GB, 21.6 GB with the fifth added, and NVIDIA's own NVFP4 build lands in the same 24 GB band. Runtimes earn the full 25, with released llama.cpp, Ollama, vLLM, SGLang, LM Studio and Transformers all running it. Builds earn 15 for NVIDIA's own FP8 and NVFP4 builds. Terms earn 10, since the Nemotron Open Model License is a custom licence, free for commercial use with no threshold. Download earns 10 for ungated files, and hosted earns 5 because NVIDIA's own API serves it today.

Step 3.7 Flash 70

Memory 5, the smallest build at 4 bits is StepFun's IQ4_XS GGUF, 104.99 GB with the 3.97 GB projector, and 108.96 GB × 1.2 = 130.8 GB, over what the reference machine can give a model and within 256 GB. The IQ3 and Q3 builds are below 4 bits and aren't read. Runtimes 20, Released vLLM 0.33.0, SGLang 0.5.23 and Transformers 5.22.2 have the step3p7 model, and llama.cpp b12242 has the architecture, with StepFun's own GGUF. Four at 5 each, and the card documents all four. Builds 15, StepFun publishes GGUF at five quantised precisions, an FP8 build and an NVFP4 build. Terms 15, Apache 2.0 in the card's metadata and its licence section, an OSI licence with nothing added. Download 10, the Hub API reports gated false for all four repositories on 10 November 2026, so the files download with no account and nothing to accept. Hosted 5, StepFun's own API serves it, per the card and our StepFun API listing, and Featherless and DeepInfra are live for it on Hugging Face. 5 + 20 + 15 + 15 + 10 + 5 = 70.

Nemotron 3 Nano 30B A3B 60

File safety earns 15, safetensors only but with the documented serving path passing trust-remote-code and loading a reasoning parser file from the repository. Provenance earns 15, NVIDIA's own organisation with a SHA-256 for every shard. Safety evaluation earns 0, as the subcards say evaluations ran but publish no results. Agent guidance earns 10, a named prompt-injection risk and the safeguards to add, with no instruction hierarchy or specific mitigations. Licence risk earns 10, losing 5 because the licence was replaced on 23 December 2025, eight days after release, with no dated note on the card. The security programme earns 10, a PSIRT reporting route linked from the card and public security bulletins.

Step 3.7 Flash 45

File safety 15, safetensors and GGUF only, no .bin, .pt or .pth files (25). Provenance 15, the files sit in the stepfun-ai organisation, which the card's StepFun platform and model page belong to (10). Safety evaluation 0, no safety results on the card. Agent guidance 0, the card has no limitations section and gives no guidance on prompt injection, instruction priority, refusals or safeguards. Licence risk 15, Apache 2.0 has no use policy, threshold or change clause. Security programme 0, no way to report a flaw on the card, and our StepFun API listing found no security.txt, bug bounty or disclosure policy. 15 + 15 + 0 + 0 + 15 + 0 = 45.

Nemotron 3 Nano 30B A3B 90

The chat template earns 20 as a Jinja file in the repository. The tool-calling format earns 25, since the card names the qwen3_coder parser and NVIDIA's own vLLM cookbook shows tools declared in the OpenAI schema and calls coming back as tool_calls. Trained tool calling earns 20, multi-step tool use in training and several calls to a turn in the template. Structured output earns 5, a claim of structured output training with no way to ask for JSON. Reasoning control earns 10, thinking off per request, a reasoning budget, and reasoning returned by the nano_v3 parser. Context earns 10 for 1M tokens with no scaling.

Step 3.7 Flash 85

Chat template 20, chat_template.jinja in the repository (5,723 characters). Tool-calling format 25, the card names the parser to use in two runtimes, step3p5 in vLLM (with auto tool choice) and in SGLang, and both ship Step parsers in a release. Trained tool calling 20, the card presents tool orchestration as a main use, with multi-turn results, and the template renders every call in a turn. Structured output 0, Nothing on the card about JSON or schema-constrained output for the weights. Reasoning control 10, the template has a reasoning_effort setting and keeps reasoning_content, and the card names the step3p5 reasoning parser for vLLM and SGLang, which returns the reasoning apart from the answer. Context 10, 256K tokens per the card, and 262,144 positions in config.json. 20 + 25 + 20 + 0 + 10 + 10 = 85.