MiniMax M3.1 Flash Is Live on Multos — A 428B-Parameter MoE at $0.15/M
MiniMax M3.1 Flash (October 19, 2026) is a multimodal open-weight MoE — 23B active of 428B parameters, 1M context, $0.15/$0.60 per million tokens. It's live on Multos Lite via Together, Fireworks, Novita, and Parasail.
MiniMax released M3.1 Flash on October 19, 2026, and it is now live on Multos Lite and above — served through Together, Fireworks, Novita, and Parasail, which means four independent providers behind every request. It is an open-weight, multimodal mixture-of-experts model that activates 23 billion parameters out of 428 billion total, with a 1M-token context and a list price of $0.15 per million input tokens and $0.60 per million output tokens.
What MiniMax M3.1 Flash Actually Is
M3.1 Flash is the flash tier of MiniMax's M3 generation, and the upgrade over M3 is broad rather than narrow: MiniMax reports significant improvements across complex agentic tasks, coding, and cybersecurity. The model accepts text, image, and video input and returns text, with a 1,048,576-token context window. MiniMax positions it for complex, multi-step co-work automation, understanding real codebases, fixing bugs, building features, and visual understanding and creation — including frontend and presentation work. Because the weights are open, the model is served by a dozen independent providers, which is why Multos can offer it on four. Pricing is $0.15/$0.60 per million tokens with cached input at $0.015; the free launch period ended October 22, 2026, and list price is now in effect.
Benchmarks: Strong and Fast, With One Honest Caveat
Artificial Analysis scores M3.1 Flash at 42.7 on its Intelligence Index, with 48.0% on Humanity's Last Exam and 53.4% on SciCode. Vals AI measured 73.9% on Vibe Code Bench v1.1, 58.8% on Finance Agent v2.1, and 67.1% on its Legal Index — business-adjacent scores that punch above the price class. It is also genuinely fast: providers on OpenRouter's network measure roughly 130-230 output tokens per second at the median. The caveat, stated plainly: the model hallucinates more than its peers on knowledge-intensive questions. Its non-hallucination rate on AA-Omniscience is 7.2%, and coverage of the launch flagged exactly this. For grounded work — coding against a real codebase, transforming content you supply — that matters little. For open-domain factual recall, keep retrieval or verification in the loop; Multos' build-time verification gates already do part of this job.
What This Means for Building on Multos
M3.1 Flash sits on the Lite tier, which means entry-plan pricing on a model with a 1M context and native vision. Four providers behind it is the part that matters operationally: if one has a bad minute, the fallback chain moves to the next without you noticing. Auto Router treats it as an efficient-tier option for high-volume work — the 23B-active MoE keeps per-token compute low, which is where the speed comes from. The practical use: everyday builds, UI and frontend generation, multimodal input, and long agentic runs where cost per completed task is the metric that matters.
Where to Find It
Find MiniMax M3.1 Flash in the MiniMax group of the model dropdown, with entries for Together, Fireworks, Novita, and Parasail. It is available on Lite and above.
Ready to Get Started?
MiniMax M3.1 Flash is live on Multos Lite and above — a 428B-parameter open-weight MoE with 1M context, native vision, and four providers behind every request. Pick it from the MiniMax group or let Auto Router route high-volume work to it.
Start Building FreeRelated Articles
Claude Opus 5.5 Is Live on Multos — Fable-Class Work at 40% Lower Cost
Anthropic released Claude Opus 5.5 today and it's immediately available on Multos Elite. It matches Claude Fable 5.1 on most tasks, costs 40% less to run than Opus 5, and is 30% faster. On Terminal-Bench 4.0 it scores 66.4% — higher than GPT-6 Astra, higher than Opus 5, higher than Fable 5.1. Here's what that actually means for building apps and games on Multos.
Read more →GPT-6 Sol and GPT-6 Luna Are Live — OpenAI Halves the Price of Its Best Models
OpenAI released GPT-6 Sol and GPT-6 Luna today, cutting prices by 50% versus GPT-5.6. Sol drops from $4/$20 to $2/$10 per million tokens. Luna drops from $0.20/$1.20 to $0.10/$0.50. Intelligence Index scores hold steady, hallucinations drop significantly, and both models are live on Multos now — Sol on Pro, Luna on Lite.
Read more →MiMo V2.6 Flash and Pro Are Now on Multos — Frontier-Grade Agents at Flash-Tier Pricing
Xiaomi released MiMo V2.6 Pro and Flash on September 21, 2026. Pro scored 46 on the Artificial Analysis Intelligence Index — the highest any open-weight model has ever scored, at $0.13 per task. Flash costs $0.14/$0.28 per million tokens and trails Pro by only 4 points on DeepSWE. Both are live on Multos Lite via DeepInfra.
Read more →