DeepSeek-V4-Flash-0731 vs Hy-Memory-30B-A3B

DeepSeek-V4-Flash-0731 38

Model card 8, the card and the DSpark card it names give architecture, parameters and context, with intended use, limitations, languages and cut-off missing. Training data 5, a general description in the report with no cut-off. Last change 10, a card edit on 1 August 2026, 93 days before 2 November, so past 90 and within 180; deepseek-recipe 0.2.0 (20 October) stays read as a library-wide release, not a change to this model's files. Responsiveness 0, DeepSeek answered none of the 61 discussions, the newest opened on 17 October. Version policy 10, the preview and DSpark copies keep their own repositories and the dated 0731 release says what it supersedes. File health 5, the files agree with each other, but reports of tool-call errors and looping and a pull request for a missing chat template wait without a reply from DeepSeek.

Hy-Memory-30B-A3B 68

Model card 8, Intended use (the write path of Mem0, MemOS and Hy-Memory), parameters (30B, about 3B active) and languages (English and Chinese, in the metadata) are there. Training data 5, A general description of the two training stages inside the harnesses, with no datasets, sizes or cut-off. Last change 20, the last commit, on 16 October 2026, added the licence and the card, 25 days before this check. Responsiveness 20, Of the 4 discussions that aren't pull requests (16 to 30 October 2026), 3 have a reply from a tencent organisation member. Version policy 5, A first release, with nothing replaced. File health 10, Config, tokenizer and template agree on the architecture, the context and the thinking levels. 8 + 5 + 20 + 20 + 5 + 10 = 68.

DeepSeek-V4-Flash-0731 70

Memory 5, because the smallest build of 4-bit or better from a runtime's own library is LM Studio's MXFP4 GGUF at 156.4 GB, which with a fifth added needs 187.7 GB, more than our 107.4 GB machine and within 256 GB. DeepSeek's own release, 166.9 GB, sits in the same band. Runtimes 25, the cap, from vLLM v0.32.0, SGLang v0.5.22, llama.cpp release b11771, Transformers 5.20.0 and LM Studio's library, with Ollama not counted since its library entry is retired with no models pushed. Builds 10, DeepSeek ships one build, the FP8 and FP4 release itself. Terms 15, MIT with nothing added. Download 10, not gated. Hosted 5, DeepSeek's API moved to V4.1 Flash, but five inference providers on Hugging Face serve this repository live (Together's mapping has gone since 21 October).

Hy-Memory-30B-A3B 85

Memory 25, Tencent's own Q4_K_M GGUF is 18.24 GB, and 18.24 GB × 1.2 = 21.9 GB, which fits 24 GB but not 16 GB. Its Q3_K_M file and the two files with experts at 1 or 2 bits are below 4-bit precision and don't count. Runtimes 20, hy_v3 model code in released vLLM (HYV3ForCausalLM in the v0.33.0 registry), SGLang v0.5.23 (hunyuan_v3.py), Transformers v5.22.2 and llama.cpp b12242, 5 each. Builds 15, Tencent publishes an FP8 repository and a GGUF repository with Q4_K_M, Q6_K and Q8_0 files in its own organisation, so quantised builds at two precisions or more. Terms 15, Apache 2.0, with only a preamble naming Tencent as copyright holder above the licence text. Download 10, gated is false for the main, FP8 and GGUF repositories on 10 November 2026, so the files download with no account or terms to accept. Hosted 0, Hugging Face lists no inference provider for the model (an empty provider mapping on 10 November 2026), and the card names no API. 25 + 20 + 15 + 15 + 10 + 0 = 85.

DeepSeek-V4-Flash-0731 50

File safety 15, safetensors only, less 10 because prompts are built by running DeepSeek's encoding_dsv4.py from the repository and the card's serving commands pass --trust-remote-code. Provenance 15, the deepseek-ai organisation and a SHA-256 on every shard. Safety evaluation 0, none published. Agent guidance 0, the encoding README's notes about the parser and the reserved developer role are usage notes, read alike for all three DeepSeek listings. Licence risk 15, MIT with no deductions. Security programme 5, for the security@deepseek.com reporting address in DeepSeek's site footer, read as for DeepSeek-V3.2, with no security.txt and no public advisories.

Hy-Memory-30B-A3B 60

File safety 25, 51 safetensors shards and no pickle or Python files. Provenance 15, files in the tencent organisation that the card's own resource list links to (10), with a SHA-256 listed for all 53 large files (5). Safety evaluation 0, no safety results on the card. Agent guidance 0, no prompt-injection guidance, instruction hierarchy, account of trained refusals or general cautions on the card. Licence risk 15, Apache 2.0 with no use policy, no field-of-use limit, no threshold and no clause letting Tencent change the terms. Security programme 5, Tencent's Security Response Center takes vulnerability reports company-wide with published reward rules, as the Hy3 listing records (5). 25 + 15 + 0 + 0 + 15 + 5 = 60.

DeepSeek-V4-Flash-0731 75

Chat template 10, there is no Jinja template and the format is given as Python to run. Tool-calling format 10, DeepSeek documents how tools are declared and returned and ships its own parser, but only vLLM's recipe names a runtime parser, so the 25 line isn't met. Trained tool calling 20, several invoke elements in one call block. Structured output 15, DeepSeek's deepseek-recipe library documents JSON Schema response formats for V4 prompts. Reasoning control 10, chat mode with no thinking and three effort levels in thinking mode, with reasoning returned as reasoning_content. Context 10, 1M tokens trained.

Hy-Memory-30B-A3B 65

Chat template 20, chat_template.jinja in the repository. Tool-calling format 10, the card doesn't mention tools. Trained tool calling 10, the template can render tool calls, but the card presents the model as a memory writer that returns memories as text or JSON and says nothing of calling tools. Structured output 5, the card's quickstart asks for a JSON object with a memories list in the system prompt and shows nothing else, no response_format and no schema, so JSON output is shown by example without a documented way to constrain it. Reasoning control 10, the card passes reasoning_effort as no_think per request through chat_template_kwargs, and the template takes high, low or no_think. Context 10, 262,144 tokens native (max_position_embeddings in config.json). 20 + 10 + 10 + 5 + 10 + 10 = 65.