LLM models

A fast and cost-effective model designed for code, agents, and complex workflows, with a 1-million-token context and significant improvements on…

DeepSeek-V4.1-Flash

4.4(8)

A multimodal model with 552B parameters, of which only 8B are active at the input stage, featuring native visual understanding and accessible via…

GLM-5.3

A long-term coding and agent model based on GLM-5.

GLM-5.3-Flash

The first natively multimodal model in the GLM-5 family, capable of observing interfaces and generating code based on what it sees, with a context…

GLM-5.2

An open-source model designed for code and long-running tasks, with a context window of up to 1 million tokens.

Mistral Large 4

5.0(1)

A powerful open-source model (1,000 billion parameters) that understands text and images and can analyze technical drawings and satellite imagery.

Gemini 3.6 Flash

Designed for intensive tasks, this model from Google’s Gemini family outperforms its predecessor in token efficiency while improving performance in…

MiMo-V2.6

4.4(7)

A series of (multimodal) LLM models able to process text and images for reasoning tasks.

Gemini 3

4.6(7)

Google's most intelligent multimodal model: advanced reasoning, deep contextual understanding, autonomous agents, visual coding, and Deep Think mode.

Kimi K2.6

4.4(11)

This open-source model takes agent-based programming even further with multi-file editing, 256K of context, and the generation of animated front-end…

Inkling

Discover Thinking Machines’s first LLM: an open, multimodal model with a context size of up to 1 million tokens.

Muse Spark

4.5(10)

The first model developed by Meta Superintelligence Labs, built from scratch in nine months and natively multimodal (text, image, voice).

Qwen 3

4.6(12)

A powerful multilingual AI model, capable of step-by-step reasoning or instantaneous response.

Muse Glimmer

4.4(9)

A 30-billion-parameter agent-based model capable of running directly on a Mac or PC equipped with a consumer-grade GPU, with no internet connection…

DeepSeek V4

Deepseek's open source LLM with 671 billion parameters specializing in mathematics and reasoning.

Kimi K2 Thinking

The new agentic and open-source model from Moonshot AI (supported by Alibaba).

Kimi.ai

4.5(4)

A powerful multimodal model for converting images into usable data.

Amazon Nova

4.5(4)

Explore a range of height multimodal AI models created by Amazon.

Kimi K2.5

4.5(6)

An open-source multimodal model with 1T parameters and a swarm agent orchestrating up to 100 sub-agents in parallel.

ERNIE 5.1

Baidu's flagship model is more cost-effective and powerful: it requires only one-third of the resources of its competitors and achieves a score of 99.

Gemma 4

Take advantage of enhanced audio and visual understanding to create comprehensive multimodal applications, ranging from image analysis to audio…

Grok 4.1

The upgraded version of Grok 4, with hallucinations reduced by 3 times, superior emotional intelligence, and creativity increased by 600 points.

GLM 5

This 745B-parameter Mixture-of-Experts model aims for complete agentic intelligence with planning, tools, web browsing, and context up to 200K…

DeepSeek-VL2

4.5(6)

A multimodal model that makes images and text much easier to understand.