Skip to content
Modelwire
Subscribe

MiniMax M3.1 Flash matches Deepseek on performance and price

Source published ·Modelwire updated

Original coverage: The Decoder ↗·How Modelwire adds context

Illustration accompanying: MiniMax M3.1 Flash shows China's AI price race has no floor in sight

The development

MiniMax's M3.1 Flash represents a significant escalation in China's competitive push toward commodity-priced AI models. The release achieves a 13-point jump in capability metrics while cutting costs in half relative to its predecessor, now matching Deepseek's V4.1 Flash on both performance and pricing. This pattern signals a structural shift in the market: Chinese vendors are systematically compressing the cost-to-capability curve faster than Western incumbents, forcing a recalibration of unit economics across the industry. For enterprises and API consumers, the immediate effect is margin pressure on inference providers, but the longer-term implication is that capability parity at sub-$0.30 per task may become table stakes rather than a competitive advantage.

Modelwire’s AI-generated summary of coverage from The Decoder.

Modelwire analysis

Analyst take

Our AI-generated reading of the wider context and the next developments to watch.

The real story isn't M3.1 Flash's specs but the timing: this launch arrives just three days after China's export restrictions on advanced models, suggesting Beijing is using open-release cadence as a counterweight to capability lockdown. The cost parity with Deepseek signals not competition between Chinese vendors but coordinated market positioning.

This follows directly from China's export control policy announced on October 16, which carved out a fast-track exemption for open-source registrations while blocking proprietary model shipments. MiniMax and Deepseek are now executing that playbook: releasing capable open weights domestically while Western incumbents (Google's Gemini 4 Argon, Meta's Muse Wonder Full from mid-October) are forced to compete on inference cost rather than capability exclusivity. The margin pressure on API providers mentioned in the summary is already visible in Meta's tripled inference costs, which makes commodity pricing from Chinese vendors a structural threat to Western provider unit economics.

If MiniMax and Deepseek maintain sub-$0.30 parity on the next benchmark cycle (typically 4-6 weeks out) while Western providers hold pricing above $0.50 for comparable tasks, that confirms the cost floor has genuinely shifted and Western API providers will need to either accept margin compression or exit inference entirely. Watch whether OpenAI or Anthropic announce cost reductions in the next 30 days; if they don't, it signals they're betting on capability moat over price competition.

This interpretation is generated from the summary above and the archive coverage cited below. Our methodology · Report an error

Coverage behind this analysis

These archive entries ground the connection in our analysis. They are ordered by source publication date, with links to our coverage and the original sources.

  1. ·The Decoder

    China restricts exports of advanced AI model weights and training data

    China's new export restrictions target advanced AI systems exceeding 10^26 FLOPs or 2 trillion parameters, blocking shipments of model weights, source code, and training methodologies to foreign buyers. The policy carves out a fast-track exemption for open-source registrations, creating a bifurcated market where proprietary Chinese AI remains locked down while open releases face minimal friction.…

    Read Modelwire coverage →Original source ↗

MentionsMiniMax · M3.1 Flash · Deepseek · V4.1 Flash · Artificial Analysis Intelligence Index · DeepSWE v1.1

MW

How this coverage is produced

Modelwire uses AI to generate summaries and context from source headlines, snippets, and selected archive coverage. Automated checks do not verify every claim, and items are not routinely reviewed by a person before publication. Zacaria Solis operates the site. Read the linked source for the full evidence and report errors through our corrections process.

Modelwire summarizes, we don’t republish. The Decoder originally reported this story as “MiniMax M3.1 Flash shows China's AI price race has no floor in sight”. The full content lives on the-decoder.com. If you’re a publisher and want a different summarization policy for your work, see our takedown page.

MiniMax M3.1 Flash matches Deepseek on performance and price · Modelwire