@AskMaximoAI

Creators of Kydanet: AI agents with their own computers and inboxes. We build AI models, APIs, and apps. | https://xitter.cf/t.co/SBOJxeV0qY

maximoai.co/platform
Joined August 2023
  • Tweets1.3K
  • Following8
  • Followers297
Three days ago we launched Maximo Atlas 1.5. Today, two security models built on it. Introducing Maximo Atlas 1.5 Cyber: 馃數 Cyber Blue, for defense. Open to everyone. 馃敶 Cyber Red, for authorized offense. By application. Same price and speed as Maximo Atlas 1.5. 馃У
Maximo AI@AskMaximoAIOct 20
Introducing Maximo Atlas 1.5. Our first sparse model. Stronger agentic coding, better UI design, and real-time 3D from a single prompt. Up to 3脳 faster than Maximo Atlas 1.4, at the same price. Available today on Maximo AI API, Maximo Syntax, MyTabulon and Kydanet.
1
6
3
8
659
Use Maximo Atlas 1.5 today: API, with $5 free credits: maximoai.co/platform Maximo Syntax CLI & Desktop: maximoai.co/syntax MyTabulon: mytabulon.com Kydanet, every plan: kydanet.com Model ID: maximo-atlas-1.5
1
4
38
Same price as Maximo Atlas 1.4: $0.20 per 1M input tokens $0.03 per 1M cached input tokens $1.00 per 1M output tokens 1M context window, 128K max output, five reasoning levels.
1
4
26
Nothing in the launch film is mocked up. Every interface and 3D scene was built by Maximo Atlas 1.5 itself, one prompt each, in the Maximo Syntax CLI: a live dashboard, a landing page, a 150,000-particle galaxy, a city at night and more.
1
4
30
Under the hood: a 552B-parameter sparse mixture of experts that activates 8B parameters on input and 16B on output. The knowledge of a very large model, at the speed of a much smaller one.
1
4
25
Built to ship code: Terminal-Bench 2.1: 89% DeepSWE v1.1: 77.2% Terminal-Bench 4.0: 42.3% And beyond code: 86.5% on BrowseComp, 75% on Toolathlon Verified, 93.7% on CyberGym.
1
4
33
Introducing Maximo Atlas 1.5. Our first sparse model. Stronger agentic coding, better UI design, and real-time 3D from a single prompt. Up to 3脳 faster than Maximo Atlas 1.4, at the same price. Available today on Maximo AI API, Maximo Syntax, MyTabulon and Kydanet.
2
3
4
5
672
Introducing Maximo Decisions 1: ask questions, get typed decisions with probabilities. Yes/no, choice and score answers for classification, routing, moderation and grading. $0.10 / 1M input tokens 路 output is free 路 1.05M context 路 200 questions per request.
3
5
106
Introducing Maximo Atlas 1.4 Built to go beyond a single command. 90% on Terminal-Bench 2.1 at max reasoning effort. Gains on 9/10 reported benchmarks. Live in Maximo AI Chat + API. $0.20 input 路 $0.03 cached 路 $1 output / 1M tokens.
1
2
4
5
868
We're introducing Maximo Atlas 1.3. A full-stack frontier agent built for terminal, code, and web work at max reasoning effort. Now live on Maximo AI API Platform at $0.20 input, $0.02 cached input, and $0.50 output per 1M tokens. maximoai.co/platform
1
3
2
5
821
Load more