Trending in AI Right Now

Most-used models (tokens/day)

Real routed usage across thousands of apps — data via OpenRouter, 2026-07-06.

1. 1.DeepSeek V4 Flash · DeepSeek748.5B
2. 2.MiniMax M3 · MiniMax661.9B
3. 3.xiaomi/mimo-v2.5658.2B
4. 4.GLM 5.2 · Z.ai561.7B
5. 5.DeepSeek V4 Pro · DeepSeek481.4B
6. 6.anthropic/claude-4.7-opus403.8B
7. 7.anthropic/claude-4.8-opus292.9B
8. 8.tencent/hy3-preview258.6B
9. 9.stepfun/step-3.7-flash234.3B
10. 10.anthropic/claude-4.6-sonnet190.9B
11. 11.GPT-5.5 · OpenAI150.7B
12. 12.Gemini 3 Flash Preview · Google148.8B

Trending open-source models

By Hugging Face trending score (community downloads + likes momentum).

1. 1.empero-ai/Qwythos-9B-Claude-Mythos-5-1M-GGUFimage-text-to-text
2. 2.zai-org/GLM-5.2text-generation
3. 3.tencent/Hy3text-generation
4. 4.baidu/Unlimited-OCRimage-text-to-text
5. 5.InternScience/Agents-A1text-generation
6. 6.google/tabfm-1.0.0-pytorchtabular-classification
7. 7.nvidia/Qwen3.6-27B-NVFP4text-generation
8. 8.deepreinforce-ai/Ornith-1.0-35B-GGUFtext-generation
9. 9.AliesTaha/fable-tracestext-generation
10. 10.yuxinlu1/gemma-4-12B-agentic-fable5-composer2.5-v2-3.5x-tau2-GGUFtext-generation
11. 11.HauhauCS/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressiveimage-text-to-text
12. 12.deepseek-ai/DeepSeek-V4-Pro-DSparktext-generation

Rising AI repos (last 14 days)

New repositories gaining stars fastest (topics: llm, ai-agents).

1. 1.synthetic-sciences/openscience★ 1,254

   The open-source AI workbench for scientific research

2. 2.amplifthq/opentag★ 876

   Open-source @agent mentions for Slack and GitHub. OpenTag routes tagged requests to Codex, Claude Code, then returns results in thread.

3. 3.oomol-lab/open-connector★ 798

   Open-source auth gateway connecting 1000+ SaaS providers to AI agents through SDK, CLI, MCP, HTTP, and OpenAPI.

4. 4.jmerelnyc/Talos★ 723

   GPU worker client for the Talos network. Pairs with your Talos account, serves open-model inference jobs over a WebSocket, and reports uptime for payouts.

5. 5.benchflow-ai/awesome-evals★ 688

   A curated, non-BS library of the best resources for building and evaluating AI agents — papers, blogs, talks, tools, benchmarks. Maintained by BenchFlow.

6. 6.uzairansaruzi/hermex★ 675

   Native iPhone app for your Hermes agent

7. 7.lycorp-jp/sim-use★ 627

   Give your AI agent eyes and hands on iOS Simulator and Android emulator/devices.

8. 8.TianhangZhuzth/Fundamental-Ava★ 529

   Build digital human beings — autonomous, collaborative, and socially intelligent agents. FNzgGxU31RWiDgLr3GvxxSa42nRntvZNSG6aBMQ1pump

9. 9.simonlin1212/Vibe-Research★ 420

   Vibe-Research: Your Personal Trading Research Agent · A股/美股/港股 的个人投研 Agent：每日复盘、资讯雷达、个股数据、板块中心、我的持仓、研究记录。Vibe-Research 把数据和功能配齐，由你自己的 AI 驱动投资研究。

10. 10.ai4s-research/open-science★ 302

    Open Science — an open AI workbench for scientists. Open-source alternative to Claude Science: local-first, model-agnostic, reproducible AI research desktop (macOS & Windows), built on Tauri + MCP + a

SDK adoption pulse (downloads/week)

Which AI stacks developers actually install — npm + PyPI weekly downloads.