AI firms prioritize infrastructure over model improvements. September 19, 2026 · By Cordelia Ashcombe The cost of running AI models continues to decline, but the focus of competition has shifted away from the models themselves. This week's developments from Zed, Anthropic, and OpenRouter highlight why the surrounding software infrastructure now holds greater importance than the models at their core. Five of the week's most-read stories centered on a single theme: transforming AI models into practical tools for users. This involves the harness - the collection of tools, interfaces, and systems that provide context, link external services, direct tasks, and validate results. While model performance still dominates headlines, companies are increasingly directing investment toward the infrastructure that enables real-world AI applications. Zed introduces Delta threaded review system. Zed introduced Delta in public beta, a system designed to replace traditional pull requests with shared discussion threads. The concept is straightforward: code reviews should not require handling between fragmented diffs and scattered comments. Delta keeps conversations directly tied to the codebase, while DeltaDB tracks changes at a granular level - down to individual edits. Nathan Sobo, Zed's CEO, framed the transition by saying the industry is currently focused on rebuilding code collaboration. His argument centers on how traditional diffs reveal only the final state of changes while obscuring the reasoning behind them. Delta addresses this by maintaining a continuous thread of discussion alongside the code. Internally, Zed's team implemented 570 modifications to Delta's main branch without submitting a single pull request, a claim supported by the company's repository data. While the public Zed editor continues to accept pull requests, the direction is clear. GitHub's commit surge highlights collaboration demand. In contrast, GitHub reported 2.9 billion monthly commits in August, up from 1.4 billion in April. Even accounting for an eight-hour service disruption that month, the figures demonstrate Git's enduring dominance in developer workflows. Yet the demand for modernization is rising. Competitors such as Cursor's Origin and GitLab's Project Switch are also rethinking collaboration tools to accommodate AI-assisted development. Anthropic adopted a different strategy by merging its Claude Chat and Cowork interfaces into a single system. This eliminates the need for users to determine upfront which tool best suits a given task. The company said people used both tools and found the decision-making process frustrating. The update is initially available to Pro and Max subscribers, with broader access planned for later. OpenRouter offers Conversion Table-based routing service. This shift reflects broader economic pressures. Companies are prioritizing the layers surrounding AI models over the models themselves. For example, OpenRouter recently made its US in-region routing service generally available for business clients. The platform decrypts, processes, and delivers requests within U.S. data centers - or rejects them if compliance requirements cannot be met. Open models dominate usage, closed systems retain revenue. Open-weight models now account for 60% of OpenRouter's U.S.-based token consumption, with Chinese models like DeepSeek V4 Pro, Kimi K3, and GLM 5.2 driving significant demand. However, the country of origin and processing location remain distinct considerations. OpenRouter's value lies in granting companies control over the latter - a critical factor as 77% of businesses now evaluate an AI solution's country of origin when selecting vendors, according to a Deloitte survey. Stripe's reported $8 billion acquisition of OpenRouter shows the strategic importance of this infrastructure layer. Economic trends support this focus: Vercel's September AI Gateway Production Index shows open-weight models now handle 56% of token volume on its platform, up from 7% in December 2025. Despite this usage, they generate only 14% of estimated spending, while closed models such as Anthropic's Claude dominate revenue despite lower adoption rates. Teams processing over 10 million tokens monthly experienced a 7.6% reduction in per-token costs in August, following a 2.9% decline in July. The downward trend in inference pricing accelerates, but the true expense lies in the harness, the connectors, authentication systems, and observability tools required for scalable AI deployment. Real-SWE benchmark shows agent success rates. A benchmark from Specific Labs, called Real-SWE, assessed coding agents on real-world tasks drawn from private repositories. The top-performing setup, a Claude Fable 5.1 model accessed through Claude Code, succeeded in 38.8% of cases. GPT-6 Astra, when used via Codex CLI, achieved 33.8%, while Gemini 3.8 Flash through its CLI reached 31.2%. No configuration exceeded 40% success. Abhilash Rao Mesala, a Meta data engineer, outlined a practical solution in a recent guide: LLM response caching. By reusing responses when requests, context, permissions, and underlying data remain unchanged, teams can achieve substantial cost reductions. His example begins with 1 million monthly calls priced at $0.006 each, totaling $6,000. With a 60% cache hit rate and $150 in additional embedding and vector-store expenses, the total drops to $2,550, a 57.5% savings. This approach leverages a principle predating transformer models. Vercel's AI Gateway Production Index for September exposed a deeper economic divide. While open-weight models now account for 56% of all requests on the platform, closed models still capture 86% of estimated spending. This discrepancy reveals that enterprises are willing to pay premiums for consistency, even when cheaper alternatives exist. Anthropic's Claude alone represents 64% of estimated spending among closed models, despite GPT-6 Astra, launched just 12 days before the report, accounting for only 7.7% of the total. The cost of inference itself continues to fall. This marks the third consecutive monthly reduction, with Vercel reporting a 23.2% year-over-year decrease in average token pricing. The trend accelerates the push toward harness solutions, tools that manage, optimize, and secure AI interactions rather than the models themselves. Boris Renski, CEO of Apelogic, summarized this shift in a recent interview: "Companies are buying the harness around inference now, and Zed, OpenRouter, and Anthropic all spent this week selling it." The Cloud Native Computing Foundation (CNCF) reinforced this perspective, noting that enterprises now treat AI integration as an infrastructure challenge rather than a model-centric concern. The Specific Labs Real-SWE benchmark further illuminated the harness challenge: even the highest-performing agents struggle with real-world tasks. The test, conducted on private codebases, showed that Claude Fable 5.1, the top performer, succeeded in only 38.8% of attempts. GPT-6 Astra and Gemini 3.8 Flash trailed further, with none surpassing 40%

Zed launches Delta public beta, betting agent threads can replace GitHub pull requests. Zed released a public beta of Delta, a collaboration tool built around AI-agent "threads" and a new DeltaDB versioning layer meant to replace GitHub pull requests. Verified pipeline | Sources: 4 | Publisher: signed | Contributor: signed | Hash: 01e8873cff | View Overview. Zed, the company behind the open-source Zed code editor, released the public beta of Delta on September 16, a collaboration tool built around AI coding agents that replaces GitHub's pull-request review model with shared, persistent "threads," according to Zed's announcement and The New Stack. Delta had been running in a private beta since August, when Zed first revealed the product, per The New Stack. What The Machine Herald know. * Threads instead of pull requests. In Delta, a developer hands a coding task to an agent, keeps discussing and refining the work in an ongoing conversation, and can invite teammates directly into that same thread rather than packaging the result into a pull request, according to Zed's blog post. Reviewers can join the original thread or create a dedicated review subthread that gets "its own isolated copy of the parent thread's worktrees," letting them test changes without disturbing the original work, per Zed. * DeltaDB underpins the system. Delta is built on DeltaDB, which "extends Git's content-based versioning with incremental versions based on deltas," recording edits and conversation messages between commits instead of waiting for work to be packaged into a single change, according to Zed. Delta still works with existing Git repositories, and developers can keep using branches, commits and remotes as before, as reported by The New Stack. * Zed has already gone all-in internally. The company disabled pull requests on Delta's own repository and now develops the product "entirely within Delta," according to Zed. Since making that switch, 33 Zed developers have landed 570 changes to their main branch without a single pull request, per Zed and The New Stack. Zed's public, open-source editor repository will "remain on GitHub for now," the company said, because it's where its outside contributor community reports issues and submits changes, according to Zed. "We're moving more thoughtfully with Zed's public repo because we have hundreds of monthly contributors who depend on that workflow," Zed co-founder and CEO Nathan Sobo told The New Stack. "GitHub has an established social component that will take longer to replace, and we're not going to strand contributors to prove a point," according to The New Stack. * Availability and pricing. Delta is available as a desktop app for macOS, Linux and Windows, plus a browser version, and remains free during the public beta, with paid individual and team plans "to follow," per The New Stack. "There will always be a free version of Delta," the company said in its announcement, according to Zed. * Company background. Zed was founded in 2021 by veterans of GitHub's Atom editor team, including Sobo, who spent nine years there, building a high-performance, multiplayer code editor in Rust, according to The New Stack. Zed raised a $32 million round led by Sequoia Capital in August 2025, when it first teased DeltaDB, per The New Stack. * The reasoning behind Delta. Sobo argues the pull request has been strained by the sheer volume of code agents now generate. "Since GitHub introduced pull requests over 15 years ago, they've become the standard way to ask teammates to review changes to your codebase," Sobo wrote in Zed's announcement. "But with agents generating so much code, the diffs we're asking each other to review have mushroomed," according to Zed. He argues threads preserve the reasoning behind a change instead of leaving reviewers to reconstruct it after the fact: "I believe the thread will replace the commit or branch as the fundamental unit of software development," Sobo said, per The New Stack. * GitHub's growing strain. Delta arrives as GitHub grapples with rapid growth in coding-agent activity. GitHub's monthly commit volume rose from roughly 1 billion across all of 2025 to 1.4 billion a month by April, then to 2.9 billion commits a month by August, according to The New Stack. GitHub also suffered a nearly eight-hour outage on August 17 that disrupted Actions, pull requests, APIs, Git operations, Webhooks and Copilot, with error rates reaching "about 20% across its web experience and API traffic" at the height of the incident, according to InfoWorld. * Not the only contender. Zed is one of several companies rethinking Git-era collaboration tools for AI agents. Cursor, now owned by SpaceX, launched its own code-hosting platform, Origin, in August, combining Git hosting, pull requests and coding agents while still treating existing GitHub repositories as the source of truth, as previously reported by The Machine Herald and per The New Stack. GitLab is separately developing a "next-generation source-code management" project called Project Switch, currently in private beta, according to The New Stack. What The Machine Herald don't know. * Sobo said Zed is "a few months away" from leaving GitHub behind for its own internal use, but gave no firm date, according to The New Stack. Zed's public, open-source repository will keep using GitHub and pull requests, with no announced timeline for when or whether that changes. * Zed has not said when Git storage will move fully into DeltaDB or when CI and release workflows will migrate away from GitHub, per Zed's blog post. * Pricing for Delta's paid individual and team plans, promised to follow the free public beta, has not been disclosed.