Coframe

AI platform that autonomously generates, tests, and improves websites, landing pages, and ads for conversion optimization.

Website: https://www.coframe.com

Cover Block

From the public record

Field	Value
Name	Coframe
Tagline	AI platform that autonomously generates, tests, and improves websites, landing pages, and ads for conversion optimization. [PR Newswire, November 2026]
Headquarters	San Francisco, California [Crunchbase]
Founded	2023 [Crunchbase]
Stage	Series A [PR Newswire, November 2026]
Business Model	SaaS
Industry	Other
Technology	AI / Machine Learning
Geography	North America
Growth Profile	Venture Scale
Founding Team	Co-Founders (2), Josh Payne and Pavlo Razumovskyi [Business Insider, November 2026] [Dev Curation, November 2026]
Funding Label	$10M+
Total Disclosed Funding	$35,000,000 [PR Newswire, November 2026]

The Short Version

PUBLIC Coframe builds software that autonomously generates, tests, and improves websites, landing pages, and ads, and it merits investor attention now because its November 2026 financing brought total disclosed funding to $35 million while public coverage tied the product to a broader push toward continuously adaptive digital experiences [PR Newswire, November 2026] [Business Insider, November 2026]. Founded in 2023 and based in San Francisco, the company was started by Josh Payne and Pavlo Razumovskyi, with Payne presented in public coverage as an AI researcher and entrepreneur and Razumovskyi identified as Coframe's co-founder and engineering lead [PR Newswire, November 2026] [Business Insider, November 2026] [Dev Curation, November 2026].

The product pitch is straightforward, even if the technical claims still lean heavily on company and investor framing: Coframe says its platform replaces a manual conversion-optimization workflow across design, engineering, growth, and analytics by autonomously ideating, building, testing, and analyzing experience variants [PR Newswire, November 2026] [Antler, November 2026]. That positioning is differentiated less by a new ad or CMS surface than by the claim that optimization becomes continuous and model-driven, a thesis the company and its backers describe as a move from static pages toward a "living internet" [Business Insider, November 2026] [Antler, November 2026].

On team quality, the public record is stronger on founder narrative than on operating metrics. Payne's background spans Autograph and earlier startup roles cited across Business Insider, podcast material, and older TechCrunch coverage, while Razumovskyi's profile points to prior technical leadership before Coframe [Business Insider, November 2026] [Apple Podcasts, July 2025] [TechCrunch, January 2014] [Crunchbase].

Coframe's business model is SaaS, and the financing base now includes Khosla Ventures, Hanabi Capital, Antler, and Nat Friedman, with the November 2026 Series A disclosed at $25 million and total disclosed funding at $35 million [PR Newswire, November 2026] [Antler, November 2026] [Crunchbase]. What matters over the next 12 to 18 months is whether the company can convert a strong narrative into independently verifiable traction: repeatable enterprise adoption, clearer proof that customer revenue impact extends beyond company-reported figures, and evidence that its autonomous optimization workflow can hold up across more channels and larger go-to-market teams [Business Insider, November 2026] [Khosla Ventures, November 2026].

Single-source, plausible -- Core company, founder, and funding facts are corroborated by multiple public sources, but product performance and differentiation claims rely materially on company, investor, and profile-driven disclosures.

Taxonomy Snapshot

Axis	Value
Stage	Series A
Business Model	SaaS
Industry / Vertical	Other
Technology Type	AI / Machine Learning
Geography	North America
Growth Profile	Venture Scale
Founding Team	Co-Founders (2)
Funding	$10M+ (total disclosed ~$35,000,000)

The Company in Brief

PUBLIC

Coframe appears in the public record as a San Francisco company founded in 2023, with Josh Payne and Pavlo Razumovskyi listed as co-founders on its company profiles [Crunchbase]. The plain-English pitch is consistent across those profiles: the company is building software that automatically creates, tests, and improves websites, landing pages, and advertisements, with the stated aim of making digital experiences adapt over time rather than stay fixed after launch [Crunchbase].

The milestone trail that can be verified from company-profile data is still short, which is typical for a young venture-backed software company. Crunchbase shows Coframe as founded in 2023 and based in San Francisco, then later records a November 2026 Series A financing that brought total disclosed funding to $35 million [Crunchbase]. That chronology supports a simple read of the business so far: an early product build period beginning in 2023, followed by outside financing once the company had established enough market interest to raise a larger institutional round [Crunchbase].

Single-source, plausible -- This section relies on a single public company-profile source, Crunchbase, for founding date, headquarters, founders, company description, and financing chronology.

What They Have Built

MIXED

Coframe is positioned around a simple claim that is easy to state and harder to verify in depth: it wants digital experiences to keep changing after launch, rather than remaining fixed until a human team updates them. Public materials describe the product as an AI platform that continuously generates, tests, analyzes, and improves websites, landing pages, and advertisements [PR Newswire, November 2026]. Antler's investment note is more specific about the workflow substitution, saying the system autonomously ideates, designs, builds, tests, and analyzes experience variants in place of the usual handoff chain across growth, design, engineering, analytics, and marketing teams [Antler, November 2026]. Business Insider frames the same thesis as a "living internet," meaning interfaces that adapt continuously to user behavior and business goals rather than being treated as static pages [Business Insider, November 2026].

The technical description that is publicly available remains high level, but it is directionally consistent across sources. Crunchbase says Coframe trains multimodal models that understand users and business goals so interfaces can improve themselves over time [Crunchbase]. Separate public signals suggest the company has also built internal developer tooling around that ambition: a project called Coffee is described on GitHub as an AI layer for frontend development, and a DeepMind engineer's LinkedIn post says Coffee was first built internally and that roughly 80% of Coframe's own frontend is AI-generated [GitHub, retrieved 2026] [LinkedIn]. Those statements help explain how Coframe may compress iteration cycles, but they do not amount to an external product demo, and the underlying stack beyond AI-assisted frontend workflows is not disclosed in the reviewed sources. An adjacent distribution signal is the reported work with OpenAI on ads and landing-page optimization for ChatGPT's ads offering, although the scope appears to be a commercial collaboration rather than a broadly documented platform integration [Unite.AI] [Dealroom News].

Unconfirmed -- This section relies heavily on company announcement language, investor materials, platform profiles, and secondary reporting, with limited independent technical verification.

Market Size and Demand

Market context

PUBLIC The market matters now because Coframe is trying to sit at the intersection of three budgets that already exist, digital advertising, website optimization, and enterprise software for experimentation, and the public record supports the intersection more clearly than it supports any standalone market size claim for the company itself [PR Newswire, November 2026] [Antler, November 2026] [Business Insider, November 2026]. The evidence base here is thin on direct TAM, SAM, and SOM figures, so the cleaner approach is to frame the addressable market through adjacent categories that buyers already fund, rather than present a precise sizing number that the available sources do not substantiate.

Coframe's product description places it inside conversion optimization and digital experience tooling, but also adjacent to performance marketing software because it continuously generates and tests ads, landing pages, and websites rather than serving only one surface [PR Newswire, November 2026]. Antler's description is useful here because it defines the workflow being replaced, growth, design, engineering, analytics, and marketing coordination, which implies Coframe is competing for spend that is usually fragmented across people, agencies, testing tools, CMS layers, and paid media experimentation budgets [Antler, November 2026]. That is commercially meaningful even without a verified market map, since software that collapses several manual steps can draw from multiple budget owners if the ROI case is legible.

A cautious TAM view, then, is that Coframe is pursuing an analogous market made up of digital experience platforms, experimentation software, and ad creative optimization tools rather than a novel budget line item. Business Insider's reporting that the company positions itself around a "living internet" thesis helps explain the strategic ambition, but it remains a company-framed narrative rather than a third-party market definition [Business Insider, November 2026]. For public investors or private market readers, that distinction matters, because category language can travel faster than procurement behavior.

Analogous market lens	What the sources support	Relevance to Coframe
Conversion optimization software	Coframe says it continuously generates, tests, and improves websites, landing pages, and ads [PR Newswire, November 2026]	Directly aligned with performance and web conversion budgets
Digital experience and web operations	The product spans websites and landing pages, not just ad creative [PR Newswire, November 2026]	Pulls the company closer to CMS, experimentation, and digital product tooling
Marketing automation and creative testing	The workflow described includes ideation, design, testing, and analysis for experience variants [Antler, November 2026]	Positions Coframe against manual campaign optimization and agency work
Enterprise AI application software	The company trains multimodal models to understand users and business goals, according to Crunchbase [Crunchbase]	Frames Coframe as an AI application layer sold on business outcomes rather than model access

The practical takeaway is that Coframe's opportunity likely expands or contracts with the willingness of growth and digital teams to consolidate fragmented tooling. That can be a large pool of spend, but the current public material does not support a precise market size figure, so the better read is budget adjacency rather than hard TAM arithmetic.

Demand drivers are easier to substantiate than market size. The strongest driver in the record is organizational friction: Antler argues that traditional conversion optimization requires cross-functional coordination across growth, design, engineering, analytics, and marketing, while Coframe is built to automate that loop [Antler, November 2026]. If that framing holds in practice, the product addresses a familiar enterprise pain point, which is not simply low conversion rates but the slow speed at which teams can generate, deploy, and learn from new variants.

A second tailwind is the broader shift toward AI-native software that acts on workflows rather than merely assisting users. Crunchbase's company description says Coframe trains multimodal models that understand users and business goals, and Business Insider reports that Khosla's investment thesis centers on replacing static websites with adaptive interfaces [Crunchbase] [Business Insider, November 2026]. Those are company and investor claims, so they should not be mistaken for independent proof of adoption. Still, they do line up with a wider enterprise preference for software that can close the loop between analysis and execution, particularly in revenue-facing functions.

A third tailwind is that buyers already measure digital channels in economic terms. Business Insider reports Coframe's claim that it has generated more than $300 million in incremental revenue for customers, including Dropbox, BCG, Block, Backcountry, and Sur La Table [Business Insider, November 2026]. That figure is company-reported and unverified in the available public record, but the form of the claim is still notable: the product is being sold against incremental revenue, not against softer engagement metrics. Markets tied to attributed revenue often support faster budget approvals than tools sold on productivity alone, assuming the measurement framework is trusted.

The adjacent and substitute markets are broad enough to create both opportunity and friction. On one side, Coframe can benefit from demand flowing out of web analytics, A/B testing, personalization, content management, growth marketing, and creative automation, because its pitch crosses those boundaries [PR Newswire, November 2026] [Antler, November 2026]. On the other, that same breadth means buyers may compare it with incumbents they already use in experimentation or digital experience stacks, even if Coframe's autonomous workflow is more expansive than a standard testing tool.

There is also a substitute dynamic worth watching around internal tooling. Public references to Coffee, an internal front-end engineering tool later shared publicly, and claims that a large portion of Coframe's own front end is AI-generated suggest part of the market may attempt to build adjacent capabilities in-house rather than purchase a dedicated optimization platform [GitHub - Coframe/coffee] [Sebastien Lohier - Google DeepMind | LinkedIn] [Coframe on X, December 2023]. That does not weaken the category by itself, but it does imply the strongest commercial wedge may be orchestration, measurement, and continuous deployment, not simple code generation.

Regulatory and macro forces are less explicit in the sources, but a few are visible by implication. Because Coframe's system is described as understanding users and business goals and optimizing digital experiences continuously, deployment likely sits near the same privacy, consent, and measurement constraints that shape modern advertising and website personalization more broadly [Crunchbase] [PR Newswire, November 2026]. The public sources do not specify how the company handles those constraints, so there is no basis to infer compliance posture. Investors should simply recognize that any platform promising automated experience adaptation operates in a market where signal loss, platform policy changes, and privacy rules can affect model performance and attribution.

Macro conditions are somewhat more favorable than they would be for a discretionary design tool. The company appears to sell into functions tied to revenue efficiency, growth, and customer acquisition, which can remain funded even in tighter software budgets if payback is clear [Khosla Ventures, November 2026] [Business Insider, November 2026]. At the same time, enterprise AI budgets have become crowded, and buyers increasingly ask whether a product replaces existing spend or adds a new layer. Coframe's market case is strongest if it can prove consolidation and measurable lift, rather than rely on category novelty alone.

Single-source, plausible -- Market interpretation is grounded in PR Newswire, Antler, Business Insider, Crunchbase, and public recruiting materials, but no named third-party TAM report or independently verified market sizing figures were available in the reviewed sources.

Who Else Is Fighting for This

MIXED Coframe is positioning itself less as a point optimization tool and more as an autonomous layer for digital conversion work, which places it between established experimentation platforms, creative automation vendors, and in-house growth engineering teams [PR Newswire, November 2026] [Antler, November 2026].

The immediate competitive map is easiest to read by workflow. On one side are incumbents in experimentation and digital experience infrastructure, the systems most enterprises already use to test pages, route traffic, and measure outcomes. The available source set does not name those vendors directly, but Coframe's own framing is clear enough: it wants to replace a manual process that typically spans growth, design, engineering, analytics, and marketing with one AI system that ideates, builds, tests, and analyzes variants [Antler, November 2026]. That implies competition not only with software budgets, but also with internal teams and agency workflows that still own landing pages, paid creative, and web experimentation.

A second cluster is adjacent creative and model-layer tooling. Coframe's public materials describe continuous optimization across websites, landing pages, and ads, and outside coverage ties the company to OpenAI's ChatGPT ads effort for combined ad and landing-page optimization [PR Newswire, November 2026] [Unite.AI] [Dealroom News]. If that relationship is commercially meaningful, the company is competing for a broader share of performance marketing workflow than a standard A/B testing vendor. The substitute here is not only another startup, but any stack where media teams generate creative in one tool, ship traffic to static pages built elsewhere, and analyze conversion later in separate analytics products.

The company's clearest edge today appears to be talent density and ambition of scope, not distribution. Public reporting ties Coframe to Khosla Ventures, Hanabi Capital, Antler, and Nat Friedman, and recruiting materials describe a team of competitive programmers, former founders, and AI researchers with backgrounds in open-source code-generation projects [PR Newswire, November 2026] [Antler, November 2026] [Khosla Ventures, November 2026]. Product evidence also points to unusual internal automation depth: an internal front-end tool called Coffee was built to accelerate UI iteration, and company-linked sources say roughly 80% of Coframe's frontend is AI-generated [Sebastien Lohier - Google DeepMind | LinkedIn] [GitHub - Coframe/coffee] [Coframe on X, December 2023]. That edge is meaningful if faster product iteration translates into better experimentation loops for customers, but it is also perishable because model capabilities and AI-assisted development practices diffuse quickly across the sector.

The main exposure is that Coframe does not appear to own a default distribution channel. The sources name investors, product vision, and hiring activity, but they do not show a proprietary traffic source, a locked-in data network, or a large existing software footprint that would make adoption automatic [Business Insider, November 2026] [PR Newswire, November 2026] [Khosla Ventures, November 2026]. Its reported work with OpenAI could become a wedge if it yields privileged access or embedded distribution, but the public evidence stops short of establishing exclusivity or durable platform control [Unite.AI] [Dealroom News]. In practical terms, that leaves Coframe exposed to larger platform vendors that can bundle optimization into broader marketing or web stacks, and to internal product organizations that prefer to keep experimentation logic close to first-party analytics and brand governance.

The most plausible 18-month scenario is that the category separates into orchestration winners and feature vendors. Coframe is a plausible winner if autonomous optimization across ad creative and landing pages proves measurably better than fragmented workflows and if the OpenAI-linked channel develops into repeatable demand generation [Business Insider, November 2026] [Unite.AI]. The likely loser in that case is the disconnected manual workflow itself, especially for mid-market and enterprise teams that already spend across growth, design, and engineering to ship incremental tests [Antler, November 2026]. The inverse scenario is also easy to see: if autonomous page and ad optimization becomes a standard feature set inside larger experience, commerce, or ad platforms, standalone vendors without entrenched distribution will face a harder path to durable account control. On the public record, Coframe has enough product ambition and investor support to matter, but its defensibility still looks more execution-led than structurally locked in.

Single-source, plausible -- Section relies on two independent public sources for positioning and funding context, with competitive inferences built from company framing and investor materials rather than named, independently verified competitor disclosures.

Opportunity

Upside if execution holds

PUBLIC The prize here is not a better A/B testing tool, but a plausible shift in how digital interfaces are produced and improved: if Coframe can make websites, landing pages, and ads behave like continuously learning systems, it could sit in the control layer for a large share of commercial web conversion activity [PR Newswire, November 2026] [Antler, November 2026] [Business Insider, November 2026].

The headline opportunity is to become the operating system for conversion optimization across paid acquisition and owned web surfaces. That outcome is reachable, rather than purely aspirational, because the company is not pitching a narrow copy-generation point solution. Public sources consistently describe the product as a system that generates, tests, analyzes, and improves websites, landing pages, and advertisements, with the stated goal of replacing a manual workflow that usually spans growth, design, engineering, analytics, and marketing teams [PR Newswire, November 2026] [Antler, November 2026]. If that workflow truly collapses into one software layer, the budget line Coframe is attacking is larger than experimentation software alone.

The early signs also point to enterprise ambition instead of a prosumer wedge. Business Insider reported that Coframe says it has generated more than $300 million in incremental revenue for customers including Dropbox, BCG, Block, Backcountry, and Sur La Table, while investor job listings point to expansion across sales, business development, and growth functions, including EMEA roles [Business Insider, November 2026] [Khosla Ventures, November 2026]. Those claims remain materially company-supplied, so they should be handled with care. Even so, the combination of a $35 million disclosed funding base and investor support from Khosla Ventures, Hanabi Capital, Antler, and Nat Friedman gives the company time and backing to pursue a category-sized position rather than a feature sale [PR Newswire, November 2026] [Antler, November 2026].

Scenario paths to scale

PUBLIC The upside branches into a few distinct paths, and each depends on a different proof point becoming durable.

Scenario	What happens	Catalyst	Why it's plausible
Conversion OS for enterprises	Coframe becomes the default layer large companies use to generate and optimize high-intent web journeys, pulling budget from agencies, internal growth teams, and experimentation tools	A few referenceable enterprise deployments convert revenue-lift claims into repeatable procurement wins [Business Insider, November 2026]	Public descriptions position Coframe as replacing a cross-functional optimization workflow, not just adding content suggestions [Antler, November 2026] [PR Newswire, November 2026]
Unified ad-to-landing stack	Coframe wins on the idea that ads and landing pages should optimize together as one system	Deeper commercialization of its reported work with OpenAI around ChatGPT ads and linked landing-page optimization [Unite.AI] [Dealroom News]	The reported OpenAI relationship fits Coframe's core thesis that acquisition creative and destination pages should learn together, which is a differentiated product story if verified at scale [Unite.AI] [Business Insider, November 2026]
AI-built interface platform	Coframe extends from optimization into broader interface creation, using internal tooling such as Coffee to accelerate product velocity and lower deployment friction	Internal developer tooling matures into a durable product advantage in speed, iteration cost, and experimentation volume [GitHub] [Sebastien Lohier - Google DeepMind	LinkedIn]

What stands out in these scenarios is that each one expands the scope of value capture. The first scenario captures software budget tied to conversion and experimentation. The second reaches performance marketing spend, where the economic prize is larger if the product can influence both click-through and post-click conversion. The third is more indirect, but potentially powerful, because faster internal shipping can improve product margins and shorten the cycle time between idea, test, and deployment [Antler, November 2026] [GitHub].

What compounding could look like

PUBLIC The compounding mechanism, if Coframe is right, is a data-and-workflow flywheel rather than a classic consumer network effect.

Each additional campaign, landing page, or website variant gives the system more observed outcomes on what combinations of design, copy, layout, and audience context drive conversion, at least in theory [PR Newswire, November 2026] [Crunchbase]. If customers let Coframe control more of the stack, from ad to page to iteration cycle, the company could gain three reinforcing advantages: better performance data, deeper product embed, and lower customer willingness to revert to fragmented manual processes. That matters because a team that no longer needs to route every test through separate design, engineering, analytics, and marketing owners may not want to reassemble that workflow once the software is working acceptably well [Antler, November 2026].

There are a few public hints that this flywheel may already be forming. The company's materials describe autonomous generation and testing across multiple surface types, not one narrow page builder [PR Newswire, November 2026]. Hiring is not confined to research engineering either; recent listings include account executive, business development, growth marketing, and growth engineering roles, which is consistent with a company trying to convert product performance into broader go-to-market reach [Khosla Ventures, November 2026] [Built In, September 2026]. If the reported customer revenue outcomes hold up under diligence, the strongest compounding effect would be simple: proven lift creates customer trust, trust expands scope, expanded scope creates more training data and workflow dependence, and that in turn can improve results further [Business Insider, November 2026].

The size of the win

PUBLIC The upside is easiest to frame against software categories that own revenue-critical workflows, even if no perfect public comparable appears in the cited source set.

A conservative way to express the win is qualitative but still concrete: if Coframe becomes the control plane for enterprise web conversion, it would be competing for budgets that sit across experimentation software, marketing technology, agency services, and internal growth operations, not merely for seat-based design software [Antler, November 2026] [PR Newswire, November 2026]. With $35 million in disclosed funding already in place, the company has at least enough public financing to attempt that broader platform build [PR Newswire, November 2026]. A credible bull case is that Coframe becomes valuable in the way infrastructure companies do when they directly influence customer revenue outcomes rather than internal productivity alone. Any valuation translation from that premise would be a scenario, not a forecast, because the public record here does not provide verified revenue, retention, pricing, or independent market-size data.

On the evidence available, the large outcome is still best phrased as conditional. If Coframe can verify that its system consistently drives measurable incremental revenue for recognizable enterprise customers, and if it can extend that proof from websites into the linked ad-to-landing workflow it has discussed publicly, then the company could justify platform-level strategic value well beyond a standard martech multiple (scenario, not a forecast) [Business Insider, November 2026] [Unite.AI] [PR Newswire, November 2026]. The gap between product story and durable category position is still real, but the upside case is legible.

Single-source, plausible -- This section relies on one independently reported article and multiple company, investor, or profile sources. Several material upside premises, including customer revenue impact and product breadth, remain company-attributed or only partially corroborated.

Sources