Author: Wall Street Insights, Bu Shuqing
Original Title: Goldman Sachs In-Depth Report: Who Will Be the Long-Term Winner in China's AI Large Model Industry?
China's AI large models are standing at a historic inflection point. Goldman Sachs believes that the intelligence performance of China's open-source/open-weight large models is approaching that of the world's top proprietary models, and adoption scale by domestic enterprises and global SMEs is rapidly expanding, thereby forming a data flywheel effect that will further drive model iteration and upgrades.
According to Zhui Feng Trading Desk, the latest Goldman Sachs report points out that this evolution trajectory can be summarized as 'from DeepSeek's cost-efficiency moment last year to Zhipu GLM's model intelligence moment this year'. The team led by Goldman Sachs analyst Ronald Keung systematically evaluates four core questions in this 50-page report: how Chinese AI models achieve high performance at low cost, why they choose the open-source route and how to monetize, where the core addressable market is, and who will be the long-term winners.
In assessing the competitive landscape, Goldman Sachs introduced a "competitive positioning framework" based on pricing power, cost advantage, and financial strength. Based on this, it determines that in the foundational text model field, Zhipu (initially covered) and DeepSeek (unlisted) have the strongest positioning; in the multimodal field, ByteDance (unlisted) is leading. Goldman Sachs also maintains Buy ratings on MiniMax and Kuaishou.
Small but Mighty, Efficiency Wins
The core reason Chinese large models can achieve performance close to their US counterparts at significantly lower cost lies in dual breakthroughs in architectural innovation and parameter efficiency.
The Goldman Sachs report points out that the parameter scale of Chinese open-source models is generally between 200 billion and 1.6 trillion, only 2% to 10% of the world's top models, primarily due to limited access to high-end computing power. Meanwhile, innovations like Mixture of Experts (MoE) architecture and sparse attention mechanisms have reduced the proportion of actually activated parameters to total parameters to only 3% to 5%, significantly lowering training and inference costs.
At the specific model level, DeepSeek V4 Pro has 1.6 trillion parameters, Zhipu GLM5.2 has 0.7 trillion, and MiniMax M3 has 0.4 trillion.
Goldman Sachs attributes the recent leap in Chinese models' coding capabilities to the synergistic effects of data curation, reinforcement learning fine-tuning, and other factors. On June 27th, DeepSeek launched the speculative decoding framework DSpark, already deployed in the online services of V4-Flash and V4 Pro, boosting per-user generation speed by 60% to 85% (V4-Flash) and 57% to 78% (V4 Pro) without altering model weights or output quality.
Meituan's LongCat 2.0 released on June 30th is viewed by Goldman Sachs as a major milestone in the localization of China's AI infrastructure—this is China's first fully open-source 1.6-trillion-parameter MoE model trained and deployed entirely on 50,000 domestic compute cards. Goldman Sachs believes this proves the feasibility of a localized hardware stack during the compute-intensive pre-training phase, holding profound significance for China's AI models to reduce dependence on foreign high-end chips.
A Two-Tiered Market, the Strong Get Stronger
Goldman Sachs describes the Chinese AI model market as an emerging "two-tiered structure" and identifies two ARR-maximizing quadrants.
In the high-end market, top models represented by Zhipu GLM5.2 and Alibaba's Qwen3.7 Max are priced at approximately $1 per million tokens, five times that of low-end models, with an estimated inference gross margin of about 10% to 20% (Goldman Sachs estimate). In comparison, top US models are priced at $4 to $8 per million tokens. Chinese high-end models are only 10% to 25% of that price, but can still maintain positive gross margins due to their lower activated parameter ratio.
In the low-end market, models targeting agent tasks are priced as low as $0.06 to $0.2 per million tokens, opening up markets for price-sensitive global SMEs and individual users. MiniMax derives 60% to 70% of its revenue from overseas. Notably, DeepSeek has announced the introduction of peak/off-peak pricing for its V4 series from mid-July, with peak rates being twice the off-peak rate, resulting in a blended price of approximately $0.35 per million tokens (V4 Pro) and $0.12 per million tokens (V4 Flash).
Goldman Sachs predicts that API and subscription revenue from Chinese AI models will grow from an estimated 35 billion RMB in 2026 to 879 billion RMB in 2030, corresponding to daily token consumption increasing from 350 trillion to 4.6 quadrillion tokens, a roughly 25-fold increase.
Open-Source Strategy: Broad Penetration, Monetization Paths Await Upgrade
The Goldman Sachs report details the strategic logic behind the prevalent open-source/open-weight approach among Chinese AI models and its monetization limitations.
The core advantages of the open-source strategy are deployment flexibility and community ecosystem. Alibaba's Qwen series, DeepSeek, Zhipu GLM, and MiniMax M3 all adopt open-source or open-weight approaches, with ByteDance's Seed model being a major exception, taking a fully closed, proprietary route. The open-source model allows flexible deployment both within and outside mainland China and accelerates iteration through community feedback.
However, Goldman Sachs points out that the ARR numbers disclosed by open-source model companies likely severely underestimate the actual deployment scale and revenue potential. Taking Zhipu as an example, its ARR target for the end of 2026 is $1 billion, but the actual global deployment volume of GLM5.2 will far exceed the token volume and revenue from Zhipu's own API channels—Alibaba Cloud's Bailian MaaS platform can directly host the GLM5.2 open-source model without paying any fees to Zhipu.
Goldman Sachs expects the industry to gradually migrate from pure open-source (MIT license, completely free) to an "open-weight + community license" model—where commercial use requires signing a revenue-sharing agreement with the model company. The MiniMax M series has already adopted this model. Goldman Sachs believes this shift will significantly improve the unit economics for AI model companies, as they can benefit from revenue-sharing agreements with platforms like AWS Bedrock and Alibaba Cloud Bailian without bearing the inference compute costs themselves.
From "Token Maximization" to ROI Priority
Goldman Sachs characterizes international market expansion as the most important upside for Chinese AI models, especially in non-US markets.
Goldman Sachs' US research team estimates that by 2030, agent AI will drive a 24-fold increase in global token consumption to 120 quintillion tokens per month, with enterprise agents contributing a 55-fold increase and consumer agents a 12-fold increase. In global (non-China) markets, Chinese AI models have already achieved significant token share growth leveraging performance improvements and price advantages.
The Goldman Sachs report notes that the AI usage paradigm for global enterprises is undergoing a fundamental shift from "token maximization" to "ROI priority." The former prevailed from late 2025 to early 2026, where companies equated high token consumption with organizational productivity; the latter focuses more on clear task boundaries, daily active agent count, backend process automation, and tangible output. Data from a Jellyfish AI Engineering Trends study shows that heavy AI users in enterprises consume 10x the tokens but only achieve a 2x increase in output.
On the channel front, Alphabet's Gemini Enterprise Agent Platform and Amazon's AWS Bedrock already offer hosting services for Chinese AI models like DeepSeek, MiniMax, Moonshot, GLM, and Qwen. According to The Wall Street Journal, Microsoft's CEO recently stated that Microsoft is considering hosting a version of DeepSeek on Copilot as an optional low-cost model, emphasizing that if DeepSeek were hosted, the model would run within Microsoft's cloud ecosystem, ensuring customer data remains within Azure.
Who Are the Long-Term Winners?
Goldman Sachs constructed a three-dimensional competitive positioning framework, using quantitative metrics to assess each player's long-term winning probability. The core formula is: ARR Scale × Gross Margin Advantage + Financial Strength.
The pricing power dimension examines release speed (compared to previous generation and peer models), LMArena Arena ranking (based on large-scale blind user evaluation), and blended price per million tokens.
The cost advantage dimension examines throughput (tokens per second), cache hit rate, activated parameter ratio, and inference gross margin. The financial strength dimension examines cash on hand, net cash as a percentage of total assets, and valuation multiples.
In the foundational text model field, Goldman Sachs identifies Zhipu (initially covered, Neutral rating, target valuation $110 billion) and DeepSeek (unlisted) as having the strongest positioning, with both showing outstanding performance in pricing power and cost advantage. The aggregate implied valuation of independent AI model companies exceeds $200 billion.
In the multimodal/video generation field, ByteDance leads with Seedance. According to LatePost and 36Kr reports, Seedance has a gross margin as high as 70%, and its ARR run rate already exceeds $2 billion. Kuaishou's Kling and MiniMax's Hailuo/upcoming H3 model are also viewed favorably by Goldman Sachs, expected to benefit in the second half of 2026 from functional breakthroughs in video generation and LLM integration, as well as healthy pricing driven by supply tightness.
Goldman Sachs maintains a Buy rating on MiniMax with a target price of HK$860, citing its M3 model's position in the ARR-maximizing quadrant of high token volume and attractive pricing, and its current valuation of only 13x 2026 year-end ARR, representing a significant discount compared to valuation multiples of Chinese and global peers, with a risk-reward skewed to the upside.
