MiniMax vs smallest.ai

MiniMax M3.1 Flash: Redefining Agent Performance and Cost

Today, we officially roll out MiniMax M3.1 Flash. As a next-generation Flash model, M3.1 Flash delivers a generational leap in agentic tasks, coding, Computer Use and cybersecurity, while making online inference nearly three times faster and reducing usage costs by 70%.

AI News Roundup: Model Launches, Policy, and Infrastructure

This is a roundup of AI news from mid-October 2026, covering multiple model launches, including DeepSeek V4.1 Flash 1017, MiniMax M3.1 Flash, and Mercury 3, as well as policy updates and infrastructure developments. DeepSeek's new model is highlighted as a value leader in open-weight models, while MiniMax's release is noted for its performance claims. The article also covers Anthropic's Public Benefit Fund and a draft federal bill, US-China AI sovereignty issues, and new inference infrastructure like vLLM v0.32.0. It includes a section on top tweets by engagement.

• DeepSeek V4.1 Flash 1017 is a 552B-total / 16B-active MoE model with a 1M-token context window and MIT license.
• MiniMax M3.1 Flash is priced at $0.15/$0.60 per million input/output tokens.
• Anthropic commits ~20% of equity to the public good via a new Public Benefit Fund.

MiniMax M3.1 Flash Released

MiniMax has added a new model, MiniMax M3.1 Flash, to its LLM lineup, as indicated by the addition of a link to the model's blog post in the navigation.

Smallest.ai Lands $13M Series A for Voice AI

Smallest.ai, a voice AI startup founded in 2023, announced a $13 million Series A funding round on July 31, 2026, led by Seligman Ventures with participation from Sierra Ventures and 3one4 Capital. Additional backers included Better Capital, Upsparks Capital, Schema Ventures, Tiny VC, DeVC and Mission Street Capital. Total funding now exceeds $21 million. The capital will advance a real-time speech-to-speech system designed to reduce latency in enterprise voice interactions, targeting financial services, healthcare and contact centers. The round followed the launch of Lightning V3, V3.1 and V3.2, adding 15-language support, production voice cloning from five seconds of audio, and instruction-following controls. The company highlighted its Hydra asynchronous architecture, which processes listening, reasoning, acting and responding in parallel. Smallest.ai also referenced a May hardware partnership with Tenstorrent for cost-efficient inference.

• Smallest.ai raised $13M in Series A funding led by Seligman Ventures.
• Total funding surpassed $21 million after the round closed on July 30, 2026.
• The round included Sierra Ventures, 3one4 Capital, Better Capital, Upsparks Capital, Schema Ventures, Tiny VC, DeVC and Mission Street Capital.

Smallest.ai raises $13M for humanlike voice AI

Smallest.ai, a startup founded in late 2024, raised $13 million in a Series A round to develop a small, voice-specific AI model intended for real-time conversational agents that sound human. The round was led by Seligman Ventures with participation from Sierra Ventures and 3one4 Capital, bringing total funding to over $21 million. Smallest.ai's approach uses a lightweight real-time voice model for immediate interaction and falls back to larger foundational LLMs when needed. The company targets enterprise customer support use cases—handling accents, multiple languages, and noisy environments—and counts RingCentral and Truecaller among existing customers. Competitors named include ElevenLabs, Cartesia and regional player Sarvam. Sudarshan Kamath is founder and CEO.

• Smallest.ai raised $13 million in a Series A round led by Seligman Ventures.
• Series A participants include Sierra Ventures and 3one4 Capital.
• The funding brings Smallest.ai's total financing to over $21 million.