Devstral Small 2 vs GLM-5.3

Devstral Small 2 90

Memory 25, the smallest build at 4 bits from a runtime's own library is Ollama's devstral-small-2 at Q4_K_M, 15 GB, and 15 GB × 1.2 = 18 GB, within 24 GB. Mistral's own FP8 files are 25.79 GB. Runtimes 25, Released vLLM 0.33.0, SGLang 0.5.23 and Transformers 5.22.2 have the mistral3 model, llama.cpp b12242 has the architecture with a GGUF from its own organisation, and Ollama's library has devstral-small-2. Builds 10, one quantised precision from Mistral, the FP8 weights it releases. Terms 15, Apache 2.0 in the card's metadata and text, an OSI licence with nothing added. Download 10, the Hub API reports gated false on 10 November 2026, so the files download with no account and nothing to accept. Hosted 5, Mistral's own API serves Devstral Small 2: the card says the model is available through its API and points there when a local runtime underperforms. 25 + 25 + 10 + 15 + 10 + 5 = 90.

GLM-5.3 55

Memory earns nothing, since Z.ai's smallest build is the 755.6 GB FP8 one, about 907 GB with the fifth added, and Ollama only lists a cloud entry. Four runtimes run it in a released version, llama.cpp, vLLM, SGLang and Transformers, for 20, while Ollama and LM Studio have no local build. Z.ai publishes one quantised build, FP8, for 10, since the BF16 repository is the full precision. The GLM-5.3 License is a custom licence that is free for commercial use below a revenue threshold, for 10. The files download with no account or gate, for 10, and Z.ai's own API plus eight live providers on Hugging Face serve it, for 5.

Devstral Small 2 46

Model card 11, Intended use, the architecture, the parameters and the context are on the card. Training data 0, Nothing on the card about the training data. Last change 10, the last commit, on 15 July 2026, updated the files to match mistral-common, 118 days before this check and within 180. Responsiveness 10, Of the 23 most recent discussions that aren't pull requests (9 December 2025 to 6 November 2026), 11 have a reply from a mistralai member, which is some. Version policy 5, Devstral Small 1.1 stays in its own repository, which this card links as the predecessor, with no dated note on the older card that we read. File health 10, Config, template, tokenizer and card agree, and the June and July commits fixed stray tokens and brought the files in line with mistral-common. 11 + 0 + 10 + 10 + 5 + 10 = 46.

GLM-5.3 58

The card covers intended use, languages and context and misses limitations, architecture, parameters and cut-off, for 8. Training data is 5, a description of the GLM-5 corpus in the cited report with no cut-off. The last change, a chat template speed-up on 14 October in the FP8 repository and on 15 October in the BF16 one, is within 30 days on 2 November, for 20. Responsiveness is 5, as zai-org answered 2 of the 16 discussions that aren't pull requests, the last on 8 September, which is a few, and no discussion has been opened since 10 October. Version policy is the full 10, with each GLM-5 release in its own repository and a dated blog saying what changed from GLM-5.2. File health is the full 10, with files and card agreeing and no broken-file report waiting by title.

Devstral Small 2 60

File safety 25, safetensors only, no .bin, .pt or .pth files and no Python in the repository. Provenance 15, files in Mistral's verified mistralai organisation, which Mistral's site links to per our Mistral Large 4 listing (10). Safety evaluation 0, no safety results on the card. Agent guidance 0, the card has no limitations or safety section, and nothing on prompt injection, instruction priority or what the model refuses. Licence risk 15, Apache 2.0 has no use policy, threshold or change clause. Security programme 5, mistral.ai's security.txt gives a security contact and a submission policy, as read for the Mistral Large 4 listing (5). 25 + 15 + 0 + 0 + 15 + 5 = 60.

GLM-5.3 50

File safety is full, with safetensors only, no Python in the repository and no remote code needed in Transformers 5.12.0 or later. Provenance is full, files in the zai-org organisation Z.ai links to with a SHA-256 for every shard. Safety evaluation is 0, because the blog mentions safety evaluation and hardening but publishes no results, and agent guidance is 0, with nothing on refusals, prompt injection or cautions. Licence risk is 10, losing 5 for the security review that model-as-a-service operators above US$10 billion of revenue must pass. The security programme is 0, with no security.txt, no SECURITY.md and no advisories.

Devstral Small 2 65

Chat template 20, chat_template.jinja in the repository (11,362 characters), with the tokenizer's tekken file for mistral-common. Tool-calling format 25, the card documents tools in the OpenAI function format with an example, names the mistral tool-call parser with auto tool choice for vLLM, shows the same for SGLang, and ships test prompts with the expected tool-call output. Trained tool calling 10, the card says the model excels at using tools, and its tests cover one tool call and calls made one at a time in sequence. Structured output 0, Nothing on the card about JSON or schema-constrained answers for this model. Reasoning control 0, no reasoning mode is documented, and a June 2026 commit removed stray thinking tokens from the files. Context 10, 256K tokens per the card, served at 262,144 in its vLLM command. 20 + 25 + 10 + 0 + 0 + 10 = 65.

GLM-5.3 70

A Jinja chat template sits in the repository, for 20. The tool-calling format is 10, since it is defined only by the template and the glm47 parser is named by vLLM and SGLang rather than by Z.ai. The template takes several calls in one turn, for 20. Structured output is 0, with nothing on the card and JSON documented only for the hosted API. Reasoning control is 10, a per-request budget through reasoning_effort with the reasoning in a think block, though thinking can't be turned off. Context is 10, 1,048,576 tokens native with no rope scaling.