Editorial illustration for MiniMax's M2.5 costs 1/20 of Claude Opus, covers 30% of HQ tasks
MiniMax M2 Slashes AI Costs with Open-Source Coding Model
MiniMax's M2.5 costs 1/20 of Claude Opus, covers 30% of HQ tasks
MiniMax’s latest release, the open‑source M2.5 and its lighter variant M2.5 Lightning, is drawing attention because it claims near‑state‑of‑the‑art performance while costing roughly one‑twentieth of Anthropic’s Claude Opus 4.6. That price gap alone makes the model a tempting option for firms looking to stretch limited AI budgets. Yet the headline numbers only tell part of the story; the real test is whether the model can handle everyday workloads without a drop in quality.
While the tech is impressive on paper, MiniMax says it has already put M2.5 to work inside its own headquarters. The company reports that a sizable slice of internal activity now runs on the new model, and that it is responsible for generating most of the fresh code being committed. As the MiniMax team writes in their release blog post, “we believe that M2.5 provides vi”...
In fact, MiniMax has already deployed this model into its own operations. Currently, 30% of all tasks at MiniMax HQ are completed by M2.5, and a staggering 80% of their newly committed code is generated by M2.5! As the MiniMax team writes in their release blog post, "we believe that M2.5 provides virtually limitless possibilities for the development and operation of agents in the economy." Technology: sparse power and the CISPO breakthrough The secret to M2.5's efficiency lies in its Mixture of Experts (MoE) architecture. Rather than running all of its 230 billion parameters for every single word it generates, the model only "activates" 10 billion.
Is a model that costs one‑twentieth of Claude Opus really affordable for most users? MiniMax says its M2.5 and M2.5 Lightning approach state‑of‑the‑art performance while keeping price low. The company announced the models as open source, yet the weights, code and license terms have not been published, leaving the openness claim unverified.
In practice, MiniMax has already integrated M2.5 into its own workflow: roughly 30 % of tasks at its Shanghai headquarters are handled by the model, and about 80 % of newly committed code is generated by it. However, the figures refer only to internal use; external adoption rates are not provided. The cost comparison relies on Claude Opus pricing, but the exact pricing structure for M2.5 is not detailed, making the magnitude of savings difficult to assess.
While the deployment numbers suggest a functional system, it remains unclear whether the model’s capabilities will translate to broader enterprise environments. Until the weights and licensing are made public, analysts will have limited data to evaluate the true impact of MiniMax’s claim.
Further Reading
Related Reading
Browse all Anthropic news →
Open Source UK PM vows action on Grok's deepfake scandal, Starmer condemns X
Open Source GPT-5 helps mathematicians offload tedious tasks, says Timothy Gowers
Open Source India proposes licensing and royalty rules for AI training by Google, OpenAI
Related Topic Google launches AI chips with 4× boost, lands Anthropic multibillion deal
Business & Startups
Related Topic Anthropic's Claude also citing Elon Musk's Grokipedia, reports say
LLMs & Generative AI
From Archives Jeffrey Epstein Consulted Elon Musk Ally on Potential Tesla Privatization
From Archives T‑Mobile to beta test AI Live Translation for calls in 50+ languages this spring
From Archives Anthropic aims to curb costs as it launches USD 50B of data centers in NY, Texas
Research & Benchmarks
From Archives Anthropic declines to patch reported AI agent vulnerability, cites design
Market Trends
Common Questions Answered
How much of MiniMax's internal tasks are currently completed by the M2.5 model?
According to the article, 30% of all tasks at MiniMax headquarters are currently completed by the M2.5 model. Additionally, an impressive 80% of the company's newly committed code is generated by M2.5.
How does the pricing of MiniMax M2.5 compare to Claude Opus?
The M2.5 model costs approximately one-twentieth of Anthropic's Claude Opus 4.6, making it a highly cost-effective option for organizations with limited AI budgets. This significant price reduction could potentially make advanced AI more accessible to a wider range of companies and developers.
What is the current status of MiniMax M2.5's open-source claim?
While MiniMax announced the M2.5 and M2.5 Lightning models as open source, the article notes that the weights, code, and license terms have not yet been published. This leaves the claim of openness currently unverified, despite the company's proclamations.