[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-deepseek-v4-flash-0731-agent-update":3,"news-related-fbdcfd41-bb54-487a-8319-9f35dfc82be5":38},{"id":4,"title":5,"summary":6,"content":7,"original_url":8,"source_id":9,"tags":10,"translations":24,"news_slug":31,"published_at":32,"created_at":33,"modified_at":34,"is_published":35,"publish_type":36,"image_url":14,"view_count":37},"fbdcfd41-bb54-487a-8319-9f35dfc82be5","DeepSeek-V4-Flash转正：这次升级不靠换架构","DeepSeek官方更新日志显示，V4-Flash-0731已进入API公测，模型架构和规模保持不变，主要通过重新后训练提升智能体、编程和工具调用能力。多个Agent基准明显高于V4-Pro-Preview，并原生支持Responses API和Codex适配。真正值得关注的不是版本号，而是大模型竞争正在从堆参数转向后训练与工作流落地。","## DeepSeek-V4-Flash转正：这次升级不靠换架构\n\nDeepSeek在7月31日更新API日志，宣布DeepSeek-V4-Flash正式版进入公测。开发者不需要改接口，只要把模型名设置为 `deepseek-v4-flash`，就能调用最新版本。\n\n这条消息最值得注意的地方，不是又多了一个模型名字，而是官方明确说：V4-Flash-0731与此前的V4-Flash-Preview保持相同的模型架构和规模，只重新进行了后训练。换句话说，这不是一次重新训练底座、扩大参数规模的升级，而是把重点放在模型如何完成任务上。\n\n从官方公布的结果看，提升集中在智能体和代码Agent场景。V4-Flash在Terminal Bench 2.1上得分82.7，NL2Repo为54.2，Cybergym为76.7，DeepSWE为54.4，Toolathlon verified为70.3。官方还公布了Agent Last Exam、Automation Bench和DSBench等测试结果，其中DSBench-FullStack为68.7，DSBench-Hard为59.6。需要说明的是，后两项是内部测试集，不能简单和公开榜单横向比较。\n\n这组数据透露出一个变化：大模型的竞争焦点正在从“会不会回答”转向“能不能把事情做完”。终端操作、代码仓库修改、工具调用和多步骤任务，考验的不是单轮答案漂亮不漂亮，而是模型能否规划、执行、检查，再根据错误继续修正。对于Agent来说，后训练往往比单纯增加参数更直接，因为它优化的是行为路径和任务习惯。\n\nV4-Flash还原生支持Responses API，并针对Codex做了适配。这个动作的意义在于，模型不再只是一个聊天接口，而是在主动接入开发者已经使用的Agent工作流。接口兼容降低了迁移成本，Codex适配则把模型能力直接放到代码生成、终端执行和项目维护的实际链路中。\n\n当然，benchmark不能等同于真实生产力。官方测试使用了特定的Harness、最大努力等级和固定采样参数，开发者自己的代码库、工具权限、上下文长度和失败处理机制，都可能让结果发生变化。尤其是Agent任务，模型本身只是一个环节，工具编排、沙箱安全和结果验收同样重要。\n\n我的判断是，DeepSeek这次更新的信号比一次普通的版本迭代更清楚：下一轮模型竞争未必靠更大的底座取胜，而可能靠更好的后训练、工具接口和任务闭环。对开发者来说，值得测试的不是它在聊天榜单上排第几，而是它能否在自己的仓库里少返工、少中断、少需要人工接管。\n\n如果一个模型不换架构，只靠后训练就能把Agent能力拉上去，那么未来真正拉开差距的，可能不是参数表，而是模型对工作流的理解深度。","https:\u002F\u002Fapi-docs.deepseek.com\u002Fupdates\u002F","4194681c-1a38-405d-a917-40e1dc2622ea",[11,15,18,21],{"id":12,"name":13,"slug":13,"description":14,"color":14},"6ad31a14-c0da-42df-81fd-564281f768db","agentic-ai",null,{"id":16,"name":17,"slug":17,"description":14,"color":14},"120fa59a-ff6f-4537-9bf5-f818df636a0e","benchmark",{"id":19,"name":20,"slug":20,"description":14,"color":14},"01598627-1ea6-4b27-a5d8-874971571a71","llm",{"id":22,"name":23,"slug":23,"description":14,"color":14},"7e89b5cc-57db-4f37-bc6d-28919a73931c","model-release",[25],{"id":26,"lang":27,"title":28,"summary":29,"content":30},"05cd22b9-d745-4946-b84b-12f36cffaee0","en","DeepSeek-V4-Flash goes GA without an architecture change","DeepSeek has moved V4-Flash-0731 into public beta. The official update says the model keeps the same architecture and size as the preview, while new post-training substantially improves agent, coding, and tool-use capabilities. Its benchmark results and native Responses API and Codex support point to a broader shift: model competition is moving from parameter scale toward post-training and reliable workflow execution.","## DeepSeek-V4-Flash Goes Public: This Upgrade Does Not Rely on a New Architecture\n\nDeepSeek announced in its July 31 API changelog that the official DeepSeek-V4-Flash release has entered public beta. Developers do not need to change the API interface; they can access the latest version by setting the model name to `deepseek-v4-flash`.\n\nThe most important detail is not the new model name, but the upgrade strategy. DeepSeek explicitly says that V4-Flash-0731 keeps the same architecture and model size as V4-Flash-Preview and was only re-post-trained. This is therefore not a case of rebuilding the base model or simply scaling up parameters. The focus is on how the model behaves while completing real tasks.\n\nThe published results show the largest gains in agent and coding scenarios. V4-Flash scores 82.7 on Terminal Bench 2.1, 54.2 on NL2Repo, 76.7 on Cybergym, 54.4 on DeepSWE, and 70.3 on Toolathlon verified. DeepSeek also reports 25.2 on Agent Last Exam, 25.1 on Automation Bench Public, 68.7 on its internal DSBench-FullStack set, and 59.6 on DSBench-Hard. The last two are internal evaluations and should not be treated as directly comparable to public leaderboards.\n\nThese numbers point to a shift in what matters for large language models. The competition is moving from whether a model can produce a convincing answer to whether it can finish a task. Terminal operations, repository edits, tool calls, and multi-step workflows require planning, execution, checking, and recovery from mistakes. For agents, post-training can be more directly valuable than simply adding parameters because it shapes behavior and task routines.\n\nV4-Flash also natively supports the Responses API and is adapted for Codex. This matters because the model is being placed into developer workflows rather than treated as a standalone chat endpoint. API compatibility lowers migration costs, while Codex support connects the model to coding, terminal execution, and project maintenance loops.\n\nBenchmarks still do not equal production productivity. DeepSeek used a specific harness, maximum effort settings, and fixed sampling parameters. A developer’s repository, tool permissions, context length, and failure-handling logic can produce very different results. In agent systems, orchestration, sandboxing, and result verification are just as important as the model itself.\n\nMy view is that this update sends a clearer signal than an ordinary version bump. The next phase of model competition may not be decided by the largest base model, but by better post-training, tool interfaces, and closed-loop task execution. Developers should test not where the model ranks on a chat leaderboard, but whether it creates fewer rework cycles, fewer interruptions, and fewer moments requiring human takeover in their own repositories.\n\nIf a model can improve its agent performance through post-training without changing its architecture, then the real advantage may no longer be found in a parameter table. It may be found in how deeply the model understands the workflow around the task.","deepseek-v4-flash-0731-agent-update","2026-07-31T08:00:00Z","2026-07-31T14:04:08.058082Z","2026-07-31T14:04:08.058093Z",true,"agent",175,{"items":39},[40,45,50,55,60,65],{"id":41,"title":42,"news_slug":43,"published_at":44},"c4ec4625-4a84-4c24-88f0-0ef1beb4f19e","Grok 4.6 发布:61 分追平 GPT-5.6 Sol,把长程 Agent 的 token 账单砍到四分之一","grok-4-6-agentic-cost-frontier","2026-08-14T19:00:00+00:00",{"id":46,"title":47,"news_slug":48,"published_at":49},"c94766df-827e-4e4e-a006-b6639ec76722","DeepSeek V4-Flash-0731 转正观察:权重不动,后训练把 Agent 分数打到 V4-Pro 之上","deepseek-v4-flash-0731-agent-benchmark-official-aug2026","2026-08-01T02:00:00+00:00",{"id":51,"title":52,"news_slug":53,"published_at":54},"3cc63477-1334-497d-80cb-90850c019101","DeepSeek-V4-Flash 转正:不靠换架构,只做后训练重新发力 Agent","deepseek-v4-flash-official-post-training-agent-0731","2026-07-31T08:00:00+00:00",{"id":56,"title":57,"news_slug":58,"published_at":59},"f26ace13-9c96-47ea-a528-b6682a22aa1e","Apodex 1.1 把推理搬进真实执行:PIVOT-RL 定位关键决策点,35B mini 开源","apodex-1-1-agentic-execution-pivot-rl","2026-08-25T14:30:00+00:00",{"id":61,"title":62,"news_slug":63,"published_at":64},"e75069c6-f15c-4ff9-8b11-404d705442e8","Upstage Solar Pro 4:把「agent 跑得稳」做成新一代闭源模型卖点","upstage-solar-pro-4-agent-reliability-closed-llm","2026-08-25T03:00:00+00:00",{"id":66,"title":67,"news_slug":68,"published_at":69},"f9cf9f03-6aca-4d29-94d3-5c6acfeaf435","匿名模型 OX Alpha 短暂登顶 OpenRouter 编码榜:研究者推测底座指向智谱 GLM-5.x","ox-alpha-stealth-openrouter-glm-5-zhipu","2026-08-24T03:00:00+00:00"]