Nanbeige4.2-3B vs ThinkingCap (Qwen3.8 27B)

Nanbeige4.2-3B 72

Model card 17, Intended use, limitations, the architecture, the parameters (3B outside the embeddings), the context and the languages (in the metadata) are on the card. Training data 5, A general description of how the fine-tuning and reinforcement learning data was built and filtered, with no sources, sizes or cut-off. Last change 15, the last commit, on 11 September 2026, updated modeling_nanbeige.py, 60 days before this check and within 90. Responsiveness 20, Of the 34 most recent discussions that aren't pull requests (21 July to 10 September 2026), 20 have a reply from a Nanbeige member, which is most. Version policy 5, Nanbeige's earlier models (Nanbeige4-3B, 4.1-3B) stay in their own repositories, as the Base, FP8, 8-bit and DSpark builds of this one do. File health 10, Config, tokenizer config, template and card agree on the architecture and the 262,144 tokens. 17 + 5 + 15 + 20 + 5 + 10 = 72.

ThinkingCap (Qwen3.8 27B) 53

Model card 8 of 20, intended use, limitations and parameters are there, but architecture, context, languages and a cut-off aren't. Training data 0, not described. Last change 20, BottleCap's GGUF build of the model changed its card on 7 October 2026, 27 days ago. Responsiveness 10, BottleCap answered 4 of 8 discussions on the main repository. Version policy 5, the earlier ThinkingCap stays in its own repository with an undated pointer to this one. File health 10, the readable metadata and builds agree and no broken-file report waits.

Nanbeige4.2-3B 85

Chat template 20, A Jinja chat template in tokenizer_config.json (7,961 characters), with enable_thinking, preserve_thinking, tools and tool_call_format switches. Tool-calling format 25, the card says passing tools switches on the tool template and recommends tool_call_format xml, with json for compatibility, and its SGLang and vLLM commands name the nanbeige tool-call parser, which released SGLang 0.5.23 has. Trained tool calling 20, the card describes fine-tuning on tool-use environments and agent scaffolds and reinforcement learning on agent tasks, and the template tells the model it may call one or more functions and renders every call in a turn. Structured output 0, Neither the card nor the files mention JSON or JSON Schema output, beyond json as a tool-call format. Reasoning control 10, the card documents enable_thinking, on by default and False per request, and preserve_thinking for earlier turns, and names the nanbeige reasoning parser, which returns the reasoning apart from the answer. Context 10, 262,144 tokens, which the card says the model supports and config.json gives as max_position_embeddings. 20 + 25 + 20 + 0 + 10 + 10 = 85.

ThinkingCap (Qwen3.8 27B) 70

Chat template 20, the same Jinja template as Qwen3.8-27B, in the repository. Tool-calling format 10 of 25, the card names the parsers but the format lives only in the template. Trained tool calling 20, several calls a turn, and BottleCap evaluated agent tasks. Structured output 0, not documented. Reasoning control 10, reasoning_effort budgets thinking per request and the card says where the reasoning comes back. Context 10, 262,144 tokens.

Nanbeige4.2-3B 85

Memory 30, the smallest build from Nanbeige is the FP8 repository (the 8-bit GPTQ one is the same size), 5.19 GB, and 5.19 GB × 1.2 = 6.2 GB, within 16 GB. bartowski's GGUF isn't read for this line. Runtimes 15, Released SGLang 0.5.23 has the model (srt/models/nanbeige.py) with its parsers, llama.cpp b12242 has the nanbeige architecture, and vLLM 0.33.0 names NanbeigeForCausalLM in its registry, served through its Transformers backend. Builds 15, Quantised builds from Nanbeige at two precisions, FP8 and 8-bit GPTQ, each in its own repository. Terms 15, Apache 2.0 in the card's metadata, an OSI licence, with no use policy or other terms on the card. Download 10, the Hub API reports gated false for the model and its FP8 and 8-bit repositories on 10 November 2026, so the files download with no account and nothing to accept. Hosted 0, Hugging Face maps no inference provider to it, and no Nanbeige or BOSS Zhipin API for it was found. 30 + 15 + 15 + 15 + 10 + 0 = 85.

ThinkingCap (Qwen3.8 27B) 75

Memory 25 of 30, BottleCap's IQ4_XS GGUF is 15.5 GB, 18.6 GB with a fifth added, which fits 24 GB. Runtimes 20, it keeps the base's qwen3_5 architecture, which llama.cpp, vLLM, SGLang and Transformers run in released versions, while Ollama's entry is the base's. Builds 15, FP8, NVFP4, MLX and GGUF from BottleCap. Terms 10, PolyForm Small Business is free commercially only below 100 people and 1,000,000 USD (2019) revenue. Download 5, the main repository is gated behind a click-through with an account and contact details, though the GGUF, NVFP4 and MLX builds aren't. Hosted 0, no API or inference provider serves it.

Nanbeige4.2-3B 55

File safety 15, safetensors only, no .bin, .pt or .pth files (25). Provenance 15, the files sit in the Nanbeige organisation (Nanbeige LLM Lab), which the lab's arXiv report for this model links, so the publisher's own account, although the Hub hasn't verified it (10). Safety evaluation 0, no safety results on the card. Agent guidance 5, General cautions: the limitations section says outputs may be inaccurate, biased or harmful. Licence risk 15, Apache 2.0 has no use policy, no threshold and no clause letting the licensor change or revoke it. Security programme 5, BOSS Zhipin runs a security response centre for reporting flaws in its products, as read for the Nanbeige4.3-33B listing (5), though nothing links it to the model. 15 + 15 + 0 + 5 + 15 + 5 = 55.

ThinkingCap (Qwen3.8 27B) 45

File safety 25, safetensors only and no code in the repository. Provenance 10 of 15, the files are in BottleCap's verified organisation, but the Hub hides the gated files' SHA-256 from the public, so checksums couldn't be confirmed. Safety evaluation 0, none published. Agent guidance 0. Licence risk 10, 5 off for the size and revenue threshold. Security programme 0, no security.txt, disclosure policy or security contact.