MiniMax’s ARR Surges Past $800M, Analysts Bullish on AI Growth

MiniMax’s ARR Surges Past $800M, Analysts Bullish on AI Growth

TREE NEWS reports: On August 26, MiniMax released its first-half 2026 earnings, revealing a remarkable acceleration in growth and narrowing losses. According to trading desk sources, both UBS and Jefferies promptly issued research notes maintaining ‘Buy’ ratings, highlighting that ARR (Annual Recurring Revenue) elasticity exceeded expectations.

Revenue Acceleration: Open Platform as the Biggest Engine

MiniMax’s total revenue for H1 2026 reached $116.6 million, up 283% year-over-year. Breaking down by segment:

• Open Platform & AI Enterprise Services: Revenue hit $73.9 million, up 703% YoY, now representing 63% of total revenue (up from 30% in H1 2025). Growth drivers include increased paying users and enterprise clients, higher API call volumes, and rapid adoption of token packages.
• AI-Native Products: Revenue reached $42.6 million, up 101% YoY, benefiting from improved user engagement, stronger willingness to pay, and continued monetization of products like Hailuo AI.

Jefferies further analyzed the geographic breakdown: Mainland China revenue surged 434% YoY to $46 million, accounting for 39% of total; other regions grew 224% YoY to $71 million, representing 61%. As of June 30, 2026, MiniMax served over 230 countries and regions globally.

Gross Margin Pressure, but Improvement Expected in H2

H1 2026 gross margin was 17.9%, up 5.8 percentage points YoY but below Q4 2025 and full-year levels. Jefferies analysts Thomas Chong and Zoey Zong noted three drag factors:

• One-time compensation to customers in June
• M3 model’s early-stage instability requiring time to stabilize, with cost optimization (e.g., throughput enhancement) still in progress
• Promotional activities for token packages

However, management remains optimistic about H2. Jefferies quoted management: ‘Gross margin will improve in the second half, with further upside potential in 2027.’ On the cost side, UBS noted R&D expenses grew 139% YoY but fell to 255% of revenue (down 154 percentage points YoY). Sales and distribution expenses decreased 18% YoY, dropping to 23% of revenue (down 85 percentage points), reflecting a shift to organic growth strategies.

ARR Exceeds Expectations; July Token Consumption 20x January Levels

MiniMax’s August ARR surpassed $800 million, well above Jefferies’ prior forecast of $580 million. Monthly ARR drivers:

• Late June: M3 model token consumption rose steadily, with TPS (tokens per second) improving from 20 to 100.
• July: Rapid growth in text model usage boosted ARR.
• August: H3 model release contributed incremental ARR.

July token consumption reached 20 times January 2026 levels. In terms of ARR mix, B2B contributed about 80%, versus only 30% a year ago. Enterprise customers and developers grew about 10x since end of last year, now exceeding 2 million, particularly concentrated in agent-related tasks.

Upcoming Model Roadmap

Jefferies detailed MiniMax’s near-term model roadmap, including three upcoming models:

• M3.1: Further enhancing pre-training and post-training to prepare for next-generation large-scale models.
• M3 Pro: With nearly 3 trillion parameters, upgrading architecture, inference efficiency, and latency.
• H3.1: Improving model intelligence based on H3 user feedback.

Management emphasized that the core of model competition is not between internet giants and AI labs, but rather model intelligence, unit cost efficiency, and commercialization capability. On compute, MiniMax stated it has sufficient self-built infrastructure, cooperates with cloud providers, and has begun using domestic chips, with further contribution expected in Q4.

Analyst Ratings and Market Impact

UBS maintained a ‘Buy’ rating with a 12-month target price of HK$500 (current price HK$303), implying 65% upside, using a Price/ARR valuation method. Jefferies also maintained ‘Buy’ and raised its target price to HK$533 (from HK$506), implying ~78% upside from the current price of HK$299.60. Jefferies’ base case target is HK$533, bull case HK$639, and bear case HK$213.

UBS noted that due to limited consensus data for H1 2026, investors will focus on management’s commentary on latest ARR trends, end-2026 targets, model iteration roadmap, and AI investment outlook during the earnings call.

Market Implications

MiniMax’s strong performance signals robust demand for AI infrastructure and applications. For investors, this reinforces the thesis that AI monetization is accelerating, particularly in enterprise and B2B segments. The ARR surge and narrowing losses suggest that AI companies can achieve scale and improve profitability faster than expected. This could positively impact sentiment for AI-related equities, including cloud providers, chipmakers, and software firms. However, gross margin pressure highlights ongoing cost challenges, especially in model training and inference. The upcoming model releases (M3.1, M3 Pro, H3.1) will be critical to watch for sustained growth and margin expansion.

Key Takeaways for Investors

• MiniMax’s ARR growth demonstrates strong product-market fit in AI, especially in B2B applications.
• Gross margin improvement in H2 is a key catalyst to monitor; successful cost optimization could drive significant earnings upside.
• The shift to organic growth (lower sales spend) suggests improving unit economics.
• Geographic diversification (61% non-China revenue) reduces regulatory and geopolitical risk.
• Analyst targets imply substantial upside, but be mindful of high valuation multiples based on ARR.