Inteligência de modelos

Acompanhe os modelos que estão moldando a AI moderna.

Informações sobre modelos, datas de lançamento opcionais, especificações, recursos oficiais, projetos no GitHub e repositórios no Hugging Face, organizados por família de modelos e fornecedor.

35Famílias de modelos
30Lançamentos verificados
20Fornecedores cadastrados

Modelos mais recentes

Google DeepMind

Gemini

Google DeepMind’s Gemini models. Gemini 4 Carbon (November 2026) is the most capable model, for deep debugging, architecture, proofs and security audits; Gemini 4 Argon (September 2026) is Google’s recommended default for everyday coding and knowledge work. Both share a 1M-token window and the same per-token pricing, available through the Gemini API, Google AI Studio and Gemini Enterprise.

Somente API Texto / Multimodal
Anthropic

Claude

Anthropic’s Claude models. The current lineup is Claude Fable 5.2 for the hardest reasoning and agent work, Opus 5.5 as the recommended default for coding and knowledge work, Sonnet 5.5 for speed and intelligence, and Haiku 5.5 for high-volume, latency-sensitive tasks. All take text and image input, return text, and have a 1M-token context window.

Somente API Texto / Multimodal
TypeSafe AI

Jev

TypeSafe AI decision model for typed choices, scores and probabilities. No generated prose; input-token billing applies. Evaluate accuracy on your own workload.

Somente API Texto
Mistral AI

Mistral

Mistral model family from Mistral AI.

API + pesos Texto / Multimodal
DeepSeek

DeepSeek R

DeepSeek R model family from DeepSeek.

API + pesos Texto
DeepSeek

DeepSeek V

DeepSeek V model family from DeepSeek.

API + pesos Texto

Famílias de modelos

35 resultados

Google DeepMind · Lançado nov 2026

Gemini

Google DeepMind’s Gemini models. Gemini 4 Carbon (November 2026) is the most capable model, for deep debugging, architecture, proofs and security audits; Gemini 4 Argon (September 2026) is Google’s recommended default for everyday coding and knowledge work. Both share a 1M-token window and the same per-token pricing, available through the Gemini API, Google AI Studio and Gemini Enterprise.

Somente API Texto / Multimodal
Anthropic · Lançado nov 2026

Claude

Anthropic’s Claude models. The current lineup is Claude Fable 5.2 for the hardest reasoning and agent work, Opus 5.5 as the recommended default for coding and knowledge work, Sonnet 5.5 for speed and intelligence, and Haiku 5.5 for high-volume, latency-sensitive tasks. All take text and image input, return text, and have a 1M-token context window.

Somente API Texto / Multimodal
OpenAI · Anunciado jul 2026

GPT

GPT model family from OpenAI.

API + pesos Texto / Multimodal
Google DeepMind · Anunciado jun 2026

Gemma

Gemma model family from Google DeepMind.

Pesos abertos Texto / Multimodal
DeepSeek · Anunciado abr 2026

DeepSeek V

DeepSeek V model family from DeepSeek.

API + pesos Texto
Alibaba Qwen · Anunciado set 2025

Qwen

Qwen model family from Alibaba Qwen.

API + pesos Texto

Amazon Nova model family from Amazon AGI.

Somente API Texto / Multimodal / Imagem / Vídeo

Aya model family from Cohere.

Pesos abertos Texto / Multimodal

CogView model family from Z.ai / GLM.

Pesos abertos Imagem

Command model family from Cohere.

Somente API Texto / Multimodal

Cosmos model family from NVIDIA.

Pesos abertos Multimodal / Vídeo
DeepSeek · Data não fornecida

DeepSeek R

DeepSeek R model family from DeepSeek.

API + pesos Texto

ERNIE model family from Baidu.

API + pesos Texto / Multimodal

GLM model family from Z.ai / GLM.

API + pesos Texto / Multimodal

Grok model family from xAI.

API + pesos Texto / Multimodal / Imagem

Hunyuan model family from Tencent Hunyuan.

API + pesos Texto / Multimodal

Jamba model family from AI21 Labs.

API + pesos Texto

Janus model family from DeepSeek.

Pesos abertos Multimodal / Imagem
TypeSafe AI · Data não fornecida

Jev

TypeSafe AI decision model for typed choices, scores and probabilities. No generated prose; input-token billing applies. Evaluate accuracy on your own workload.

Somente API Texto

Kimi-VL model family from Moonshot AI / Kimi.

Pesos abertos Multimodal / Imagem

Fornecedores de modelos

Ver tudo →

Projetos atualizados recentemente

Repositórios oficiais do GitHub e revisados pela equipe editorial.

deepseek-ai/DeepInfer

DeepSeek’s own high-performance inference framework, open-sourced for offline inference of DeepSeek-V4.1-Flash on NVIDIA SM100 GPUs and Huawei Ascend 950 NPUs.

★ 25,7K · ⑂ 3,2K
typesafe-ai/typesafe-sdk-python

Connect Python services to the TypeSafe API using typed questions and results. The repository provides the client library rather than model weights.

★ 218 · ⑂ 30
typesafe-ai/typesafe-sdk-js

Integrate TypeSafe decisions into JavaScript and TypeScript services. Answer types are inferred from questions; the package supports ESM and CommonJS.

★ 231 · ⑂ 25
typesafe-ai/skills

Official agent skills for designing applications around TypeSafe typed decisions. Development guidance complements the runtime SDKs and API.

★ 2K · ⑂ 109
QwenLM/qwen-code

An open-source AI coding agent that lives in your terminal.

★ 27,2K · ⑂ 2,9K
NVIDIA/NemoClaw

Run agents like Hermes, LangChain Deep Agents, and OpenClaw more securely inside NVIDIA OpenShell with managed inference

★ 22,2K · ⑂ 3K
openai/codex

Lightweight coding agent that runs in your terminal

★ 107,1K · ⑂ 16,3K
NVIDIA/TensorRT-LLM

TensorRT LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and supports state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. TensorRT LLM also contains components to create Python and C++ runtimes that orchestrate the inference execution in a performant way.

★ 14,4K · ⑂ 2,7K