China’s STAR Market Now Allows AI Foundation Model Companies to List Without Revenue
Published 23 June 2026
Xia Yu
On 17 June 2026, the Shanghai Stock Exchange (“SSE”) promulgated its Implementation Guidance for the Review Rules on the Fifth Set of Listing Criteria Applicable to Artificial Intelligence Large Model Enterprises under the STAR Market (“Guidance”), effective immediately. This is China’s bespoke listing rule tailored specifically for the AI large model sector, signaling an institutional attempt to compete in global capital formation for AI. Under the Guidance, high-quality AI large model enterprises that have not yet generated significant revenue or even turned a profit, but that possess core technologies, now have a clear A-share listing pathway.
Overview of the Guidance
The Guidance comprises fifteen articles, the core of which is to permit eligible large model enterprises to avail themselves of the fifth set of listing criteria of the STAR Market. These criteria do not impose rigid requirements on operating revenue or net profit, but set out eight detailed specifications around the application of the fifth set of listing criteria to large model enterprises:
1. Business Scope and “Hard‑Tech” Attributes: Article 3 of the Guidance provides that the issuer’s principal business must be “independent research and development of AI large models, model services, or model applications”. This limitation excludes entities that merely perform simple integration or application packaging, emphasizing that the issuer must possess genuine “hard‑tech” attributes, thereby ensuring the technological purity of the listed entity.
2. Demonstrable Technological Superiority: Article 4 requires the issuer to demonstrate that it possesses “technological advantages at the forefront of the industry”, and provides specific disclosure dimensions, ranging from model scale, multimodal capabilities and agent breakthroughs, to “rankings in major domestic and international large model evaluations over the past year”. For industry‑specific large model enterprises, additional disclosure is required regarding advantages in “knowledge depth and accuracy, complex reasoning capabilities, and professional task execution”. This effectively places the burden of proving technological strength entirely on the issuer, and also requires the sponsor to issue a clear opinion.
3. Milestone Achievements: Article 5 explicitly requires that “at least one large model product has been launched and achieved scaled application” by the time of filing, and that “no material adverse technological events” have occurred thereafter. The enterprise must provide quantitative evidence such as “numbers of users or deployed devices, and recent model invocation volumes”, thereby preventing mere technological vision from directly accessing the public capital markets and ensuring that the listing applicant has indeed crossed the critical bridge from technology to product.
4. Obtaining Approvals: Article 6 requires that, in the course of launching external services and commercializing large models, the enterprise shall obtain the approvals of competent state authorities in accordance with laws, administrative regulations and relevant state provisions, and carry out the necessary assessments, filings and other procedures. This is a uniquely Chinese “access‑compliance” threshold, and is a procedural requirement of vital importance for foreign practitioners seeking to understand the Chinese regulatory landscape.
5. Industry Standing and Market Validation: Article 7 requires that the issuer “stands out in the industry with a leading ranking, occupies an important position in the industrial chain, and is able to play an exemplary and leading role in the industry”, and treats investment by sophisticated professional institutional investors as an “important consideration”. This mechanism effectively leverages the professional judgement of the private equity market to assist in the review process, forming a form of “soft endorsement”.
6. Market Space: Article 8 requires that the issuer’s business or products have a clearly defined target market, with large actual or potential demand, and that they enjoy prominent competitive advantages in aspects such as “R&D progress and key performance indicators”, with ample market space and strong future growth potential. This is the regulator’s concretization of the “growth” requirement.
7. Commercialization Arrangements: Article 9 stipulates that the issuer must formulate clear commercialization arrangements for its AI large model business or products, and at the same time must not have any matters such as “clearly insufficient commercialization expectations” that could materially affect its sustainable operating capacity. This is a “dual test” – comprising both an affirmative obligation (arrangements must be in place) and a negative prohibition (no manifest insufficiency).
8. Compliance Requirements: Article 10 requires that the issuer’s business or products comply with relevant provisions on national security, data security and privacy protection, intellectual property protection, and comply with internet information content policies and market fair competition order. Where overseas business is involved, the issuer must comply with relevant security requirements concerning technology export, cross‑border data flows, and the like. This means that large model enterprises intending to expand overseas will simultaneously face, in the course of their IPO review, multiple compliance reviews under China’s Data Security Law and the Personal Information Protection Law, as well as under foreign jurisdictions such as the EU Artificial Intelligence Act and the U.S. Export Administration Regulations.
International Comparisons
Comparison with Nasdaq: Qualitative vs. Quantitative Standards. Nasdaq allows pre‑revenue companies to go public primarily through financial‑metric combinations such as market capitalization/revenue, and has historically been used extensively by biotechnology companies. China’s STAR Market fifth‑set criteria were initially designed for pre‑revenue innovative pharmaceutical companies, and this extension to the large model field is distinctive in establishing a highly customized qualitative evaluation system (e.g., evaluation rankings, model invocation volumes, participation in national tasks, etc.). This “one‑sector‑one‑policy” regulatory approach stands in marked contrast to Nasdaq’s core reliance on general, quantitative criteria.
Comparison with HKEX Chapter 18C: Similarities and Differences in Institutional Design. The Hong Kong Stock Exchange’s Chapter 18C of the Main Board Listing Rules (“Chapter 18C”) introduced in 2023 is likewise directed at specialist technology companies that are not yet commercialized or are at an early stage of commercialization, including artificial intelligence. Chapter 18C categorizes specialist technology companies into “commercialized companies” and “pre‑commercialization companies”, requiring a market capitalization at listing of not less than HK$4 billion (Equivalent to US$510.23 million) for the former and not less than HK$8 billion (Equivalent to US$1.0247 billion) for the latter. In contrast, the Guidance offers advantages in terms of review timelines and market liquidity expectations, but also carries a more pronounced policy‑orientation; meanwhile, the HKEX provides international enterprises with the familiarity of a common‑law environment and the convenience of free capital movement. In addition, on the requirement for “sophisticated independent investors”, HKEX Chapter 18C sets more concrete quantitative thresholds (typically requiring at least two to three sophisticated independent investors to hold an aggregate of 10%–20% of shares before listing with a lock‑up period), whereas the Guidance treats institutional investor participation as an “important consideration” rather than a hard threshold, thus affording greater discretion on a case‑by‑case basis.
Conclusion
The promulgation of the Guidance is not merely an expansion of IPO rules, but a proactive attempt by China’s capital markets to compete in global pricing power for frontier technologies. As a matter of institutional design, the success of the Guidance hinges on two critical factors: whether intermediary institutions can truly assume their “gatekeeper” responsibility by conducting substantive verification—rather than formalistic review—of technological leadership, milestone achievements and commercialization prospects; and whether the regulator can, while maintaining rigorous standards, appropriately disclose the review approaches of typical cases so as to provide the market with predictable benchmarks for assessment.
Overview of the Guidance
The Guidance comprises fifteen articles, the core of which is to permit eligible large model enterprises to avail themselves of the fifth set of listing criteria of the STAR Market. These criteria do not impose rigid requirements on operating revenue or net profit, but set out eight detailed specifications around the application of the fifth set of listing criteria to large model enterprises:
1. Business Scope and “Hard‑Tech” Attributes: Article 3 of the Guidance provides that the issuer’s principal business must be “independent research and development of AI large models, model services, or model applications”. This limitation excludes entities that merely perform simple integration or application packaging, emphasizing that the issuer must possess genuine “hard‑tech” attributes, thereby ensuring the technological purity of the listed entity.
2. Demonstrable Technological Superiority: Article 4 requires the issuer to demonstrate that it possesses “technological advantages at the forefront of the industry”, and provides specific disclosure dimensions, ranging from model scale, multimodal capabilities and agent breakthroughs, to “rankings in major domestic and international large model evaluations over the past year”. For industry‑specific large model enterprises, additional disclosure is required regarding advantages in “knowledge depth and accuracy, complex reasoning capabilities, and professional task execution”. This effectively places the burden of proving technological strength entirely on the issuer, and also requires the sponsor to issue a clear opinion.
3. Milestone Achievements: Article 5 explicitly requires that “at least one large model product has been launched and achieved scaled application” by the time of filing, and that “no material adverse technological events” have occurred thereafter. The enterprise must provide quantitative evidence such as “numbers of users or deployed devices, and recent model invocation volumes”, thereby preventing mere technological vision from directly accessing the public capital markets and ensuring that the listing applicant has indeed crossed the critical bridge from technology to product.
4. Obtaining Approvals: Article 6 requires that, in the course of launching external services and commercializing large models, the enterprise shall obtain the approvals of competent state authorities in accordance with laws, administrative regulations and relevant state provisions, and carry out the necessary assessments, filings and other procedures. This is a uniquely Chinese “access‑compliance” threshold, and is a procedural requirement of vital importance for foreign practitioners seeking to understand the Chinese regulatory landscape.
5. Industry Standing and Market Validation: Article 7 requires that the issuer “stands out in the industry with a leading ranking, occupies an important position in the industrial chain, and is able to play an exemplary and leading role in the industry”, and treats investment by sophisticated professional institutional investors as an “important consideration”. This mechanism effectively leverages the professional judgement of the private equity market to assist in the review process, forming a form of “soft endorsement”.
6. Market Space: Article 8 requires that the issuer’s business or products have a clearly defined target market, with large actual or potential demand, and that they enjoy prominent competitive advantages in aspects such as “R&D progress and key performance indicators”, with ample market space and strong future growth potential. This is the regulator’s concretization of the “growth” requirement.
7. Commercialization Arrangements: Article 9 stipulates that the issuer must formulate clear commercialization arrangements for its AI large model business or products, and at the same time must not have any matters such as “clearly insufficient commercialization expectations” that could materially affect its sustainable operating capacity. This is a “dual test” – comprising both an affirmative obligation (arrangements must be in place) and a negative prohibition (no manifest insufficiency).
8. Compliance Requirements: Article 10 requires that the issuer’s business or products comply with relevant provisions on national security, data security and privacy protection, intellectual property protection, and comply with internet information content policies and market fair competition order. Where overseas business is involved, the issuer must comply with relevant security requirements concerning technology export, cross‑border data flows, and the like. This means that large model enterprises intending to expand overseas will simultaneously face, in the course of their IPO review, multiple compliance reviews under China’s Data Security Law and the Personal Information Protection Law, as well as under foreign jurisdictions such as the EU Artificial Intelligence Act and the U.S. Export Administration Regulations.
International Comparisons
Comparison with Nasdaq: Qualitative vs. Quantitative Standards. Nasdaq allows pre‑revenue companies to go public primarily through financial‑metric combinations such as market capitalization/revenue, and has historically been used extensively by biotechnology companies. China’s STAR Market fifth‑set criteria were initially designed for pre‑revenue innovative pharmaceutical companies, and this extension to the large model field is distinctive in establishing a highly customized qualitative evaluation system (e.g., evaluation rankings, model invocation volumes, participation in national tasks, etc.). This “one‑sector‑one‑policy” regulatory approach stands in marked contrast to Nasdaq’s core reliance on general, quantitative criteria.
Comparison with HKEX Chapter 18C: Similarities and Differences in Institutional Design. The Hong Kong Stock Exchange’s Chapter 18C of the Main Board Listing Rules (“Chapter 18C”) introduced in 2023 is likewise directed at specialist technology companies that are not yet commercialized or are at an early stage of commercialization, including artificial intelligence. Chapter 18C categorizes specialist technology companies into “commercialized companies” and “pre‑commercialization companies”, requiring a market capitalization at listing of not less than HK$4 billion (Equivalent to US$510.23 million) for the former and not less than HK$8 billion (Equivalent to US$1.0247 billion) for the latter. In contrast, the Guidance offers advantages in terms of review timelines and market liquidity expectations, but also carries a more pronounced policy‑orientation; meanwhile, the HKEX provides international enterprises with the familiarity of a common‑law environment and the convenience of free capital movement. In addition, on the requirement for “sophisticated independent investors”, HKEX Chapter 18C sets more concrete quantitative thresholds (typically requiring at least two to three sophisticated independent investors to hold an aggregate of 10%–20% of shares before listing with a lock‑up period), whereas the Guidance treats institutional investor participation as an “important consideration” rather than a hard threshold, thus affording greater discretion on a case‑by‑case basis.
Conclusion
The promulgation of the Guidance is not merely an expansion of IPO rules, but a proactive attempt by China’s capital markets to compete in global pricing power for frontier technologies. As a matter of institutional design, the success of the Guidance hinges on two critical factors: whether intermediary institutions can truly assume their “gatekeeper” responsibility by conducting substantive verification—rather than formalistic review—of technological leadership, milestone achievements and commercialization prospects; and whether the regulator can, while maintaining rigorous standards, appropriately disclose the review approaches of typical cases so as to provide the market with predictable benchmarks for assessment.