Search results for MinimaxM3
MinimaxM3 community
One keyword maps to one global community path.
People
Not Found
Tweets including MinimaxM3
MiniMax M3.1 Pro - animated SVG with audio of Abbey Road
MiniMax M3.1 Flash vs Claude Opus 5.5 in cinematic absurdity.
Two takes on the impossible, made with Higgsfield.
MiniMax M3.1 Flash gave knitwear a storyline.
A cozy little world with a loose thread in the plot.
Made with Higgsfield.
MiniMax M3.1 Flash in 3D product motion design.
A headphone ad you can almost hear on mute.
Made with Higgsfield.
MiniMax M3.1 Flash in cinematic motion design.
A chase scene where even the typography has somewhere to be.
Made with Higgsfield.
MiniMax M3.1 Flash just dropped and the pricing is almost free.
>$0.15 input, $0.60 output, $0.015 cached per 1M tokens
>Benchmarks land close to DeepSeek V4.1 Flash across the board
>Beats Kimi K3 and GLM-5.3 on most coding and cowork tests
• Ties Opus 5 on BankerToolBench at 76.2
FRONTIER-LEVEL SCORES AT A FRACTION OF A CENT. WHY ARE YOU STILL PAYING FLAGSHIP PRICES?
Show more
MiniMax M3.1 Flash is here 🚀
$0.15 in / $0.60 out per 1M tokens
1/3 the cost of M3, 3x the speed
74% DeepSWE, 91% BabyVision
Open weights Oct 23
Cheap agent models are getting scary good. Still running DeepSeek 1017 locally until something beats it on my Sparks 💪
Show more
MiniMax M3 is genuinely impressive
it one-shot this Kerr black hole observatory. it absolutely nailed the black hole feeling.
the entire workflow:
/goal + one spec describing the physics and art direction. everything in MiniMax Code.
It shipped the entire thing in one shot from that.
Show more
🚨BREAKING: MiniMax-M3.1-Flash is now listed in MiniMax’s API docs
Key specs:
• Frontier multimodal coding capabilities
• 1M-token context, ~4.9× M2.5/M2.7’s 204.8K
• Five thinking levels: low, medium, high, xhigh and max (default)
• HTTP, Anthropic SDK and OpenAI SDK support
Important for existing users:
starting Oct 19, M Plan and Token Plan requests to M2.5 or M2.7 are automatically routed to M3.1 Flash.
Show more
🚨 MiniMax-M3.1-Flash dropped
Remember Space Bunny? It was MiniMax-M3.1-Flash all along
> 1M context window
> Frontier-level coding
> Tunable thinking depth
> $0.165 input / $0.66 output per 1M tokens
> Cache pricing: just $0.0165
It's coming as an open-source model, with open weights expected soon
MiniMax is cooking again
Show more
