开源模型正在重塑企业 AI 的成本与基础设施

简要描述Brief Description

Y Combinator 采访了 Ollama 联合创始人兼 CEO Jeffrey Morgan，探讨开放模型为何正迅速成为企业 AI 的核心。他们讨论了开放模型在成本和控制方面的优势、编程智能体的发展、本地模型、云端模型、开放模型与前沿模型在实际应用中的分工、采用这些模型所面临的安全与运营挑战，以及 Ollama 如何让开发者能够使用碎片化的模型生态。Y Combinator interviews Ollama co-founder and CEO Jeffrey Morgan about why open models are rapidly becoming central to enterprise AI. They discuss the cost and control advantages of open models, the growth of coding agents, the practical split between local, cloud, open, and frontier models, the security and operational challenges of adoption, and how Ollama aims to make a fragmented model ecosystem usable for developers.

目录Table of Contents

• 转向开放模型
  The shift to open models
• 编程智能体与激增的 Token 需求
  Coding agents and exploding token demand
• 模型发布、微调与安全性
  Model releases, fine-tuning, and safety
• 以生产级规模部署模型
  Launching models at production scale
• 新兴的智能体技术栈
  The emerging agent stack
• 企业中的开放模型与前沿模型
  Open and frontier models in the enterprise
• 本地执行与混合路由
  Local execution and hybrid routing
• 中国、美国与开放模型竞争
  China, the United States, and open-model competition
• Ollama 的商业模式与 YC 经历
  Ollama's business model and YC experience
• 为 AI 原生的软件世界构建
  Building for an AI-native software world

转向开放模型The shift to open models

主持人：Ollama 让用户可以轻松地在本地和云端运行开源 AI 模型。数百万开发者都在使用它，企业也广泛采用，因此公司能够近距离了解人们实际下载并持续使用的模型。当前最先进的技术正在发生哪些变化？Host: Ollama makes it easy to run open-source AI models locally and in the cloud. With millions of developers using it and broad enterprise adoption, the company has a close view of what people actually download and continue to use. What is changing in the state of the art?

Jeffrey Morgan：最大的转变是转向开放模型，尤其是在企业领域。用户使用的模型来自美国和中国，其中以编程代理和 AI 助手为主，例如 OpenClaw 和 Hermes。Ollama 最初是为了在 MacBook 及其他硬件上运行模型而推出的，这些硬件包括 Nvidia、AMD 和 Intel 系统。如今，其云服务被全球各地的企业大量使用，其中中国来源的模型尤为常见；美国和德国也是重要的需求来源。Jeffrey Morgan: The largest shift is toward open models, particularly in enterprise. Usage includes models originating in the United States and China, and is dominated by coding agents and AI assistants such as OpenClaw and Hermes. Ollama began as a way to run models on a MacBook and other hardware, including Nvidia, AMD, and Intel systems. Its cloud offering now sees heavy usage of Chinese-origin models by businesses worldwide, with the United States and Germany among important sources of demand.

成本无疑是开放模型能够解决的最突出的眼前问题。但企业的长期目标，是更好地掌控 AI，并能够根据自身使用场景定制模型。因此，降低成本不只是节省开支，还能让企业有机会进行试验，并在之后针对独特的工作流程定制模型。例如，据报道，AT&T 已将其 40% 的 token 消耗量转向开放模型，同时评估来自多个地区的模型。Cost is by far the largest immediate problem that open models solve. But a business's longer-term goal is better control over AI and the ability to customize models for its own use case. Lower cost is therefore not just a savings measure; it makes it possible to experiment and later tailor models to unique workflows. AT&T, for example, was reported to have shifted 40 percent of its token consumption to open models, while evaluating models from several regions.

编码代理与暴增的 Token 需求Coding agents and exploding token demand

杰弗里·摩根：编码代理是每位开发者和每位用户的令牌用量增长的重要原因。OpenClaw 的兴起以及 Hermes 代理项目的推出，让这一机会不再局限于开发者：从事金融、支持、营销或销售工作的人，都可以把复杂任务委托给开放模型，并让它长时间运行。这类代理在选择工具和检索数据时会消耗大量令牌。Jeffrey Morgan: Coding agents account for much of the growth in token use per developer and per user. The takeoff of OpenClaw and the Hermes agent project expanded that opportunity beyond developers: people working in finance, support, marketing, or sales can delegate a difficult task to an open model and let it run over a long period. Such agents consume many tokens while choosing tools and retrieving data.

Ollama Cloud 的 token 使用量图表上出现了两个显著的转折点。第一个出现在年初，当时 Kimi、GLM 和 MiniMax 等模型的能力已足以支持编码智能体。第二个转折点则伴随着 OpenClaw 的出现，它帮助非开发者将更复杂的任务交给模型处理。开源模型的上下文窗口不断扩大，从大约 128K 增至超过一百万个 token，也助力实现了这一扩展。Two inflection points stood out in Ollama Cloud's token-use graph. The first, at the start of the year, came when models such as Kimi, GLM, and MiniMax became capable enough to power coding agents. The second came with OpenClaw, which helped non-developers hand harder tasks to models. Larger context windows, growing from roughly 128K to more than a million tokens for open models, helped make that expansion possible.

结果令人瞩目。核心工作场景中的人均用量增长了数倍，而自年初以来，云端总体用量增长了约150倍。2024年和2025年，大型开放模型通常以定制微调模型的形式使用。直到今年，开放模型的开箱即用服务才开始兴起。The result was dramatic. Per-user consumption rose several-fold for core work use cases, while aggregate cloud usage increased roughly 150-fold since the beginning of the year. In 2024 and 2025, large open models were often used as custom fine-tuned models. The out-of-the-box serving of open models began taking off only this year.

模型发布、微调与安全性Model releases, fine-tuning, and safety

主持人：微调在 2024 年初引发了巨大关注，但后来似乎成了白费功夫，因为下一次版本更新就会超越定制模型。如今它又回归了吗？还是这不过是又一轮周期？Host: Fine-tuning attracted enormous interest in early 2024, then seemed like wasted effort because the next release would overtake a custom model. Is it returning, or is this merely another cycle?

杰弗里·摩根：发布节奏正在加快，尤其是在开源领域。以 DeepSeek Flash 为例，它在一个夏天里就已经迭代了三次，而过去的发布周期更接近六个月一次。这样的节奏让定制训练更难以维持，不过微调工具正在不断改进，帮助团队跟上最新进展。Jeffrey Morgan: Release cadence is accelerating, especially on the open side. DeepSeek Flash, for example, had already seen three iterations over a summer, where releases once followed something closer to a six-month cycle. That pace makes custom training harder to maintain, but the tooling for fine-tuning is improving and helps teams stay current.

安全性与安保问题是制约采用的主要因素。前沿实验室可能会放缓步伐，以解决对齐和遏制问题，而开放权重模型则会持续进步。GLM-5.3 等模型的能力为安全与治理公司带来了机遇，因为对于许多正在考虑采用开放模型的企业而言，安全问题仍然是主要障碍。如果这些顾虑能够得到解决，对欧洲和美国的许多客户来说，采用中国的模型实验室是一个可行的选择。Safety and security are a major adoption constraint. Frontier labs may slow while addressing alignment and containment, while open-weight models continue to improve. The capabilities of models such as GLM-5.3 create an opportunity for security and governance companies, because the main blocker for many enterprises considering open models is still safety. If those concerns can be addressed, using Chinese-origin model labs is a viable option for many customers in Europe and the United States.

开放模型对于经过授权的安全测试尤其有用。闭源前沿模型通常会拒绝对产品进行渗透测试的请求，而面向安全领域的开放模型则可以针对这类工作进行训练。即使是通用开放模型也保留了安全训练，但可能更善于区分正当的防御性工作与恶意意图。Open models can be especially useful for authorized security testing. Closed frontier models often refuse a request to pentest a product, while security-oriented open models can be trained for such work. Even general open models retain safety training, but may better distinguish legitimate defensive work from malicious intent.

以生产规模部署模型Launching models at production scale

主持人：作为一个重要的分发渠道，Ollama 往往会在模型正式发布前率先推出。要大规模运营这类发布，需要具备什么条件？Host: As a key distribution channel, Ollama often sees models before general release. What does operating those launches at scale require?

杰弗里·摩根：模型推出的速度越来越快，因此 Ollama 制定了一套零日发布方案。模型必须能够在用户偏好的推理引擎中运行，做到快速且准确，与参考实现保持一致，并配备合适的应用场景和开发者工具框架。例如，一款新的多模态 DeepSeek 模型就需要更新工具框架，才能有效展现其能力。Jeffrey Morgan: Models are arriving faster and faster, so Ollama developed a day-zero launch playbook. A model has to work in users' preferred inference engines, perform quickly and accurately, match the reference implementation, and be paired with the right use cases and developer harnesses. A new multimodal DeepSeek model, for example, required an updated harness to expose its capabilities effectively.

每个模型都有其独特的架构、工具调用机制和运行层面的挑战。正式发布前，必须对最终产品进行基准测试，确保其表现符合实验室的预期。Ollama 提供三项内容：一个运行框架或 SDK、模型本身（具备足够的可靠性和云端算力），以及硬件或服务提供商支持。最后一项可能涉及推理服务提供商、Nvidia 和 Apple Silicon。性能出色的模型如果运行缓慢，也无法带来良好的用户体验。Every model has distinct architecture, tool-calling mechanics, and operational challenges. The final product must be benchmarked before release to ensure that it behaves as the lab intended. Ollama packages three things: a harness or SDK, the model itself with sufficient reliability and cloud capacity, and hardware or provider support. The last part can involve inference providers, Nvidia, and Apple Silicon. A capable model that runs slowly is not a good user experience.

获取模型的时间可能只比发布早几周，而大量集成工作有时要到最后二十四小时才完成。公司用操作系统作类比来描述自己的角色：它连接硬件驱动、提供方、推理层、应用运行时和执行框架。这种组合规模庞大且难以处理，但可复用的组件有助于开发、测试并发布一个通用运行时，使众多模型能够适配众多执行框架。Access to a model may arrive only weeks before launch, and much of the integration sometimes happens in the final twenty-four hours. The company describes its role with an operating-system analogy: it connects hardware drivers, providers, the inference layer, application runtime, and harnesses. The combination is large and difficult, but reusable pieces can make it possible to develop, test, and release a common runtime that matches many models to many harnesses.

新兴的智能体技术栈The emerging agent stack

主持人：模型与应用之间的这些层能否发展成独立的公司或服务提供商？Host: Could the layers between model and application become their own companies or providers?

Jeffrey Morgan：是的。一种有用的划分方式是将这些机会分为知识、协调和执行。知识是将公司的数据和上下文连接到模型。协调是管理分发给本地和云端环境中各个子代理的请求。执行则包括沙箱，以及随着代理越来越多地在云端运行所需的计算能力。Jeffrey Morgan: Yes. One useful framing separates the opportunities into knowledge, coordination, and execution. Knowledge is connecting a company's data and context to a model. Coordination is managing requests that fan out to subagents across local and cloud environments. Execution includes sandboxes and the compute required as agents increasingly operate in the cloud.

开源历来会围绕每一层催生出各自领域内最优秀的公司。早期的云平台会将各种能力打包进 Heroku 或 Google App Engine 等产品中，但开发者最终更青睐专业化产品。前沿实验室可能会试图将用户留在托管式智能体、上下文和记忆系统中，而开源开发者则会构建替代方案。最终很可能会出现打包与拆分并存的局面。Open source historically creates best-of-breed companies around each layer. Earlier cloud platforms bundled capabilities in products such as Heroku or Google App Engine, but developers ultimately preferred specialized products. Frontier labs may seek to keep users inside managed-agent, context, and memory systems, while open-source developers will build alternatives. The likely outcome is a mix of bundling and unbundling.

由于许多服务商都能提供开放模型，Token 正变得越来越充裕。新的稀缺资源是 Token 之上的工程能力：将智能体从一项任务编排到另一项任务、存储状态、管理凭证，以及提供安全工具。某些 harness 功能可能会下沉到模型中，但存储这类有状态问题，不可能简单地成为定期重新训练的模型的一部分。这些层面仍蕴藏着巨大的机遇。Tokens are becoming abundant because many providers can serve open models. The new scarcity is the engineering required above tokens: orchestrating an agent from one task to another, storing state, managing credentials, and providing security tooling. Some harness features may move down into models, but stateful problems such as storage cannot simply become part of a periodically retrained model. Those layers remain substantial opportunities.

企业中的开放模型与前沿模型Open and frontier models in the enterprise

主持人：开放模型和前沿模型会共同发展，还是企业支出会趋于稳定的比例？Host: Do open models and frontier models grow together, or will enterprise spend reach a steady split?

Jeffrey Morgan：Ollama 预计，企业使用的 token 中，绝大多数（可能占 80% 到 90%）将用于开放模型；但由于开放模型能降低单位成本，它们可能只占总支出的 10% 到 20%。这种充裕性让企业能够提供更多使用机会，而不是对 token 实行配给。最难的任务仍将由拥有顶尖研究人员的前沿实验室承担，而许多中等难度的任务则会结合使用开放模型和闭源模型。Jeffrey Morgan: Ollama expects a supermajority of an enterprise's tokens to use open models, perhaps 80 to 90 percent, while open models may account for only 10 to 20 percent of spending because they lower unit costs. That abundance lets businesses offer more access rather than rationing tokens. The hardest tasks will still belong to frontier labs with leading researchers, and many mid-range tasks will use open and closed models together.

客户应决定路由和编排在哪里进行。前沿模型可以处理更复杂的调度或协调工作，而开放模型则负责逐项执行具体任务。Sakana AI 和 OpenRouter 的项目展示了将两者结合的价值。这就像一家律师事务所：合伙人负责最棘手的判断，并将其他工作委派给律师。云计算早已走过类似的道路，将 DynamoDB 这类专有服务与 Postgres 这类开放技术结合起来。The customer should decide where routing and orchestration occur. A frontier model may handle harder scheduling or coordination, while open models perform line-item work. Projects from Sakana AI and OpenRouter demonstrate the value of combining both. The analogy is a law firm: a partner handles the most difficult judgment and delegates other work to associates. Cloud computing already followed a similar path, mixing proprietary services such as DynamoDB with open technologies such as Postgres.

本地执行与混合路由Local execution and hybrid routing

主持人：Ollama 同时支持云端托管模型和本地模型。这两种模式将如何发展？Host: Ollama operates both cloud-hosted and local models. How will those worlds develop?

Jeffrey Morgan：答案是另一种混合方案。较简单的任务可以在本地运行，延迟更低、每个 token 的成本也更低；企业则可使用云端模型处理更困难的工作。较新的硬件越来越有能力运行参数量在 200 亿到 400 亿之间的模型，有时甚至能运行参数量高达 1280 亿的模型。基准测试显示，Qwen 3.8 38B 在编程方面的表现可接近 Opus 4.6，同时还能在配置相对普通的 MacBook 上运行。Jeffrey Morgan: The answer is another hybrid. Easier tasks can run locally with lower latency and lower per-token cost, while businesses use cloud models for harder work. Newer hardware is increasingly capable of running models in the 20-billion to 40-billion parameter range, and sometimes as large as 128 billion parameters. Benchmarks suggest that Qwen 3.8 38B can approach Opus 4.6 for coding, while remaining usable on a relatively modest MacBook configuration.

本地使用方案包含强大的美国、欧洲和中国模型选项，包括 Llama 和 Gemma。编码代理通常最能从大型云端模型中受益，因为它们必须解决复杂问题并编写代码和测试。文档处理工作流通常可以在本地高效运行，因为这类任务更为简单。路由器可以将较简单的工作交给本地模型，并将困难的工作升级交由大型云端模型处理；由于企业硬件已经到位，这还能进一步降低成本。Local usage includes strong American, European, and Chinese model choices, including Llama and Gemma. Coding agents tend to benefit most from large cloud models because they must solve difficult problems and write code and tests. Document-processing workflows can often run well locally because the task is more straightforward. A router can send easier work to local models and escalate difficult work to a large cloud model, reducing costs further because business hardware is already available.

中国、美国与开放模型竞争China, the United States, and open-model competition

Jeffrey Morgan：如今，云端编码智能体的使用主要依赖中国来源的模型，而本地模型的使用则呈现出更强的地区多样性。这种差异既反映了模型的可用性，也反映了所部署工作负载的类型。Nvidia 的 Neotron Ultra 等模型的发布表明，越来越多大型美国开源模型正在问世。Jeffrey Morgan: Cloud coding-agent consumption is predominantly Chinese-origin models today, while local-model usage is a stronger regional mix. That difference reflects both model availability and the kinds of workloads deployed. The release of models such as Nvidia's Neotron Ultra suggests that more large U.S. open models are arriving.

这场对话还探讨了开源与闭源服务提供商之间的竞争性反馈循环。前沿实验室拥有高度集成的技术栈，涵盖应用、模型、基础设施与推理、芯片以及能源。开源模型让每一层都有更多选择，但这些选择需要经过筛选和整合。模型提供商、硬件公司、推理服务商和运行框架开发者必须协同合作，为开发者提供简单可靠的体验。The conversation also considers the competitive feedback loop between open and closed providers. Frontier labs have a highly integrated stack: applications, models, infrastructure and inference, chips, and energy. Open models create more choices at each layer, but those choices need curation. Model providers, hardware companies, inference services, and harness developers must work together to give developers a simple, reliable experience.

Ollama 的商业模式和 YC 经历Ollama's business model and YC experience

主持人：Ollama 花了不少时间打造开源项目，当时其商业模式还不明朗。团队是如何考虑盈利的？Host: Ollama spent time building an open-source project before its business model was obvious. How did the team think about making money?

杰弗里·摩根：团队看到了两条可持续的发展路径：注重隐私的 AI 产品，以及借助开放模型完成高难度任务。能在本地运行很重要，但从来不是全部。公司一直在等待产品层面达到可与闭源模型相媲美的产品市场契合度，而编程智能体成为了那个契机：AI 使用量最大的领域终于可以由开放模型来满足。Jeffrey Morgan: The team saw two durable paths: privacy-focused AI products and access to open models for difficult tasks. Local execution was important but never the entire story. The company waited for product-level product-market fit comparable to closed models, and coding agents became that moment: the largest area of AI consumption could finally be served by open models.

等待也伴随着风险。团队可能会形成一种文化，把商业化始终放在次要位置——这是创始人从早期 Docker 经历中吸取的教训。更重要的是坚持贴近客户，而不是把庞大的开源用户群体视为一个没有差别的整体。团队需要与客户交流，了解他们目前在做什么，并弄清楚他们希望在六个月内实现什么。Waiting carries risks. A team can build a culture in which monetization is never a priority, a lesson the founders learned from earlier Docker experience. The more important discipline is staying close to customers rather than treating a large open-source user base as an undifferentiated mass. Teams need to meet customers, understand their current work, and learn what they want to accomplish in six months.

Morgan 表示，即使对于第二次创业的创始人来说，YC 也很有价值，因为创业之路依然孤独。同行小组以及后来建立的 YC 人脉网络，让团队有了可以倾诉问题、借鉴过往失误，并避免重蹈覆辙的人；否则，这些失败可能会危及公司。看到一支团队真正实现产品与市场契合，也能让年轻创始人切身体会到什么是高质量、如何规范发布，以及成功的软件必须达到怎样的标准。Morgan says YC was valuable even for second-time founders because building a company remains lonely. Peer groups and the later YC network gave the team people with whom to share problems, learn from past mistakes, and avoid repeating failures that might otherwise threaten a company. Seeing a team deliver real product-market fit also gives young founders a practical sense of quality, release discipline, and the standards that successful software requires.

为 AI 原生软件世界构建产品Building for an AI-native software world

Jeffrey Morgan：来自基础设施和云计算公司的经验仍然很有价值，但许多旧有假设已不再成立。认为初创公司因为建立在另一层之上就必然脆弱的观点，在 AI 领域已不那么可靠；向技术栈上层发展，可能让公司更贴近客户。传统系统工程追求确定、经过测试和验证的行为，而语言模型本质上并不完美，有时还会有意表现出非确定性。Jeffrey Morgan: Experience from infrastructure and cloud companies remains valuable, but many old assumptions no longer hold. The idea that a startup is necessarily vulnerable because it sits on top of another layer is less reliable in AI; moving up the stack can bring a company closer to the customer. Traditional systems engineering seeks deterministic, tested, validated behavior, while language models are inherently imperfect and sometimes intentionally non-deterministic.

AI 也在改变人员配置和服务设计。十年前需要大型团队完成的一些工作，如今可以用不同的方式处理，但当生成的代码或不断演进的代码中没有任何一位工程师能完全理解每一行时，公司必须学会如何运营服务。长期维护的软件仍然需要明确责任归属：团队必须考虑，产品能否为多年后仍在使用它的客户正常运行。AI also changes staffing and service design. Some work that required large teams a decade ago can now be handled differently, but companies must learn how to operate services when no individual engineer fully understands every line of generated or evolving code. Long-lived software still needs accountability: teams must ask whether a product will work for customers who continue using it years later.

核心机遇在于整合。开发者不想在支离破碎的模型、推理技术、云服务和开发工具生态中四处摸索，也不想花几个小时去了解文档中未说明的服务提供商行为。他们想要构建软件、应用和公司。OpenRouter 等服务提供丰富的模型选择和统一计费；OpenCode 等开发工具则能让开发者连接到多种模型。随着模型和服务提供商日益丰富，将这些组件整合成可靠系统，正成为稀缺且有价值的工作。The central opportunity is curation. Developers do not want to navigate a fragmented universe of models, inference technologies, cloud services, and harnesses, or spend hours discovering undocumented provider behavior. They want to build software, applications, and companies. Services such as OpenRouter offer broad model selection and unified billing; harnesses such as OpenCode can connect developers to many models. As models and providers become abundant, bringing those pieces into a dependable system becomes the scarce and valuable work.