{"schema_version":"onlylabs.public_analysis_evidence.v1","title":"Novita AI analysis evidence pack","description":"Public onlylabs evidence pack for cited agent analysis: captured pages, ranked public signals, and stored web-search provenance used by the background analysis workflow.","url":"https://onlylabs.fyi/labs/novita","json_url":"https://onlylabs.fyi/analysis/novita/evidence.json","generated_at":"2026-06-13T13:58:34.176Z","org":{"slug":"novita","name":"Novita AI","category":"neocloud","category_label":"Neocloud","dossier_url":"https://onlylabs.fyi/labs/novita"},"analysis":null,"workflow":{"version":"onlylabs-deepagents-analysis-v3","provider":null,"model":null,"agent":null,"public_pack_mode":"local-pages-and-events","live_web_fetches":false,"note":"Public evidence exports do not trigger live Exa calls; stored Exa provenance is included when analysis metadata contains it."},"stats":{"pages":28,"events":69,"web":0,"evidence":88,"signal_desks":{"hiring":3,"forks":20,"releases":12,"talking":0,"repos":25},"data_radar_lanes":null,"data_radar_matches":null,"stored_analysis_evidence":null,"stored_analysis_web":null,"stored_analysis_signal_desks":null,"stored_analysis_data_radar_lanes":null,"stored_analysis_data_radar_matches":null},"stored_web_provenance":null,"evidence":[{"ref":"P1","kind":"page","title":"novitalabs/pegaflow v0.22.9","date":"2026-06-12T07:03:32.115473+00:00","date_source":null,"source_url":"https://github.com/novitalabs/pegaflow/releases/tag/v0.22.9","signal_url":null,"signal_json_url":null,"text":"# v0.22.9\n\nRepository: novitalabs/pegaflow\n\nTag: v0.22.9\n\nPublished: 2026-06-11T11:25:52Z\n\nPrerelease: no\n\nRelease notes:\n## What's Changed\n* refactor(core): replace KVCacheRegistration with validated KVCacheLayout by @xiaguan in https://github.com/novitalabs/pegaflow/pull/340\n* feat(pd): support mtp split connector by @GentleCold in https://github.com/novitalabs/pegaflow/pull/341\n* fix(core)!: seal instance layer topology from registered layers by @xiaguan in https://github.com/novitalabs/pegaflow/pull/346\n* feat(connector): register layer-split KV caches from cache config by @feifei-111 in https://github.com/novitalabs/pegaflow/pull/343\n* chore: bump version to 0.22.9 by @xiaguan in https://github.com/novitalabs/pegaflow/pull/347\n\n**Full Changelog**: https://github.com/novitalabs/pegaflow/compare/v0.22.8...v0.22.9"},{"ref":"P2","kind":"page","title":"Solutions Engineer (AI Cloud Infrastructure)","date":"2026-06-11T04:11:06.590183+00:00","date_source":null,"source_url":"https://jobs.ashbyhq.com/novita-ai/f0534078-76a1-478d-a349-6d409dddabb4","signal_url":null,"signal_json_url":null,"text":"# Solutions Engineer (AI Cloud Infrastructure)\n\nTeam: Go-to-Market\n\nLocation: San Mateo\n\nEmployment type: FullTime\n\nWorkplace type: OnSite\n\nRemote: no\n\nPublished: 2025-08-27T03:06:52.237+00:00\n\nAbout Us:\n\nWe are a high-growth, global AI cloud infrastructure provider at the forefront of the artificial intelligence revolution. Our cutting-edge platform offers developers and enterprises powerful, scalable, and easy-to-use solutions, including Model APIs, GPU Instances, and Serverless Computing. As businesses worldwide race to integrate AI into their products, we provide the foundational engine to power their innovation.\n\nWe are building a world-class team to serve a rapidly expanding customer base. This is a unique opportunity to join a dynamic company in a hyper-growth market, where your technical expertise will directly shape customer success and drive our business forward.\n\nThe Role:\n\nAs a Solutions Engineer, you will be the primary technical leader and trusted advisor for our customers throughout their journey. You will partner closely with the sales team to bridge the gap between complex customer challenges and our technical solutions. Your mission is to establish technical credibility, showcase the power of our platform, and architect solutions that ensure our customers achieve their AI-driven business goals.\n\nWhat You'll Do:\n\n- Technical Discovery & Solution Design: Partner with Account Executives to deeply understand customer needs, technical requirements, and business objectives. Architect elegant and effective solutions leveraging our AI infra stack (Model APIs, GPU Instances, Serverless).\n\n- Product Demonstration & Proof of Concept (POC): Lead compelling, customized product demonstrations and hands-on workshops. Design, manage, and execute successful POCs, proving the value and performance of our platform in the customer's environment.\n\n- Technical Evangelism & Trusted Advisory: Articulate the value proposition of our platform to both technical and non-technical audiences, from engineers to C-level executives. Become the go-to expert for customers on AI infrastructure best practices.\n\n- Sales Enablement & Market Feedback Loop: Create and maintain technic"},{"ref":"P3","kind":"page","title":"Account Executive","date":"2026-06-11T04:11:06.543008+00:00","date_source":null,"source_url":"https://jobs.ashbyhq.com/novita-ai/6b9784ff-bc78-42b0-bf1d-a6bac48f5b17","signal_url":null,"signal_json_url":null,"text":"# Account Executive\n\nTeam: Go-to-Market\n\nLocation: San Mateo\n\nEmployment type: FullTime\n\nWorkplace type: Hybrid\n\nRemote: yes\n\nPublished: 2026-04-01T06:26:16.916+00:00\n\nThe Role\n\nAs one of our first Account Executives, you will be the tip of the spear for our Sales-Led Growth (SLG) motion. You will be responsible for driving revenue by acquiring new B2B customers—ranging from agile AI startups to mid-market enterprises.\n\nYou are a \"full-cycle\" closer who is comfortable hunting for outbound opportunities, running technical discovery calls with CTOs and AI researchers, and navigating complex deal cycles. Because we are a cross-border team, your ability to bridge the gap between US clients and our global engineering team will be critical to your success.\n\nWhat You Will Do\n\n- Full-Cycle Sales: Own the entire sales process from prospecting and lead generation to negotiation and closing.\n\n- Outbound Hunting: Proactively identify and target AI startups, research labs, and tech companies that require scalable GPU compute and LLM API solutions.\n\n- Technical Discovery: Conduct deep-dive discovery calls with technical buyers (CTOs, VP of Engineering, Lead AI Scientists) to understand their infrastructure pain points and position our solutions effectively.\n\n- Pipeline Management: Maintain a healthy pipeline of opportunities, providing accurate forecasting and CRM hygiene.\n\n- Cross-Border Collaboration: Work closely with our global product, engineering, and marketing teams (based in China and the US) to ensure customer feedback is translated into product improvements and to secure technical support for complex deals.\n\n- Market Intelligence: Act as the eyes and ears on the ground, gathering insights on competitor pricing, feature gaps, and market trends in the Silicon Valley AI ecosystem.\n\nWho You Are (Qualifications)\n\n- Experience: 3+ years of quota-carrying B2B sales experience in Cloud Infrastructure (IaaS/PaaS), Developer Tools, or AI/Machine Learning SaaS.\n\n- Technical Empathy: You don't need to write code, but you must deeply understand the AI landscape. You know the difference between training and inference, understand what an API endpoint is, and can confidently discus"},{"ref":"P4","kind":"page","title":"Chief of Staff","date":"2026-06-11T04:11:06.469312+00:00","date_source":null,"source_url":"https://jobs.ashbyhq.com/novita-ai/aa4bf0ce-f369-47d5-a989-b7af828e4833","signal_url":null,"signal_json_url":null,"text":"# Chief of Staff\n\nTeam: Operations\n\nLocation: San Mateo\n\nEmployment type: FullTime\n\nWorkplace type: OnSite\n\nRemote: no\n\nPublished: 2026-04-16T08:57:01.752+00:00\n\nWho We Are\n\nAt Novita AI, we’re on a mission to make open-source AI models accessible to everyone—no massive teams or endless resources required. Our AI cloud platform handles all the heavy lifting in AI infra for developers and enterprises so they can focus on building the future. We believe that truly democratizing AI means letting every developer experiment and create without barriers.\n\nAbout the Role:\n\nWe are seeking a highly organized, detail-oriented, and proactive Chief of Staff to serve as the operational backbone of our North American team.\n\nYou will be the \"first line of defense\" for local compliance, managing everything from vendor contracts and budget tracking to SaaS optimization and daily administration. This is a high-trust, high-impact position with a clear growth trajectory toward a Head of Operations role for the right candidate.\n\nKey Responsibilities:\n\n- Compliance & Risk Management: Act as the local compliance gatekeeper. Proactively review sales contracts and vendor agreements before execution. Identify and resolve historical contract gaps to ensure robust risk mitigation.\n\n- Finance & Budget Tracking: Monitor the operational budget, track spending efficiency, and provide early warnings for budget overruns. Handle local tax, government correspondence, and administrative compliance loops.\n\n- Internal Efficiency & IT/SaaS Management: Audit and manage local SaaS subscriptions (e.g., Gusto, CRM tools) to optimize costs and utilization. Ensure data privacy and security standards are met in all contract and invoice management processes.\n\n- General Administration: Oversee daily office operations, support local HR functions, and provide comprehensive logistical support to ensure the frontline team can focus on business growth.\n\nQualifications:\n\n- Education: Bachelor’s or Master’s degree. Backgrounds in STEM, Data Analytics, Business Administration, Project Management, or Pre-Law are highly preferred. (Recent graduates with strong potential are welcome to apply).\n\n- Language: Fluent in Manda"},{"ref":"P5","kind":"page","title":"novitalabs/litelama repository metadata","date":"2026-06-11T04:08:54.527752+00:00","date_source":null,"source_url":"https://github.com/novitalabs/litelama","signal_url":null,"signal_json_url":null,"text":"# novitalabs/litelama\n\nDescription: lightweight LAMA inference wrapper\n\nLanguage: Python\n\nStars: 27\n\nForks: 3\n\nOpen issues: 0\n\nCreated: 2023-09-25T08:30:03Z\n\nPushed: 2023-09-28T04:02:49Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# Lite Lama - A lightweight LAMA inference wrapper\n\n```python\nfrom litelama import LiteLama\nimport requests\nfrom PIL import Image\nfrom io import BytesIO\n\ndef download_image(url):\nresponse = requests.get(url)\nreturn Image.open(BytesIO(response.content)).convert(\"RGB\")\n\nimg_url = \"https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo.png\"\nmask_url = \"https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo_mask.png\"\n\nlama = LiteLama()\nlama.to(\"cuda:0\")\ninit_image = download_image(img_url).resize((512, 512))\nmask_image = download_image(mask_url).resize((512, 512))\n\nlama.predict(init_image, mask_image).save(\"result.png\")\n```"},{"ref":"P6","kind":"page","title":"novitalabs/golang-sdk repository metadata","date":"2026-06-11T04:08:54.214686+00:00","date_source":null,"source_url":"https://github.com/novitalabs/golang-sdk","signal_url":null,"signal_json_url":null,"text":"# novitalabs/golang-sdk\n\nDescription: Golang SDK for Novita AI API (Txt2Img, Img2Img, ControlNet, VAE, LoRA)\n\nLanguage: Go\n\nLicense: MIT\n\nStars: 4\n\nForks: 3\n\nOpen issues: 0\n\nCreated: 2023-09-26T14:11:28Z\n\nPushed: 2023-11-21T04:10:37Z\n\nDefault branch: main\n\nFork: no\n\nArchived: yes\n\nREADME:\n# Novita AI Golang SDK\n\nThis SDK is based on the official [API documentation](https://docs.novita.ai/).\n\n**Join our discord server for help**\n\n[![](https://dcbadge.vercel.app/api/server/Mqx7nWYzDF)](https://discord.gg/Mqx7nWYzDF)\n\n## Installation\n\n```bash\ngo get -u github.com/novitalabs/golang-sdk\n```\n\n## Quick Start\n\n**Get api key refer to [https://novita.ai/get-started/](https://novita.ai/get-started/)**\n\n```golang\npackage main\n\nimport (\n\"context\"\n\"fmt\"\n\"time\"\n\n\"github.com/novitalabs/golang-sdk/request\"\n\"github.com/novitalabs/golang-sdk/types\"\n)\n\nfunc main() {\n// Get your API key refer to https://novita.ai/get-started/ .\nconst apiKey = \"Your-API-Key\"\nclient, err := request.NewClient(apiKey)\nif err != nil {\nfmt.Printf(\"new client failed, %v\\n\", err)\nreturn\n}\nctx, cancel := context.WithTimeout(context.Background(), time.Minute*3)\ndefer cancel()\ntxt2ImgReq := types.NewTxt2ImgRequest(\"a dog flying in the sky\", \"\", \"AnythingV5_v5PrtRE.safetensors\")\nres, err := client.SyncTxt2img(ctx, txt2ImgReq,\nrequest.WithSaveImage(\"out\", 0777, func(taskId string, fileIndex int, fileName string) string {\nreturn \"test_txt2img_sync.png\"\n}))\nif err != nil {\nfmt.Printf(\"generate image failed, %v\\n\", err)\nreturn\n}\nfor _, s3Url := range res.Data.Imgs {\nfmt.Printf(\"generate image url: %v\\n\", s3Url)\n}\n}\n```\n\n## Examples\n\n### Txt2Img with LoRA\n\nRefer to [./example/lora/main.go](./example/lora/main.go)\n\n### Model Search\n\nRefer to [./example/model_search/main.go](./example/model_search/main.go)\n\n### ControlNet QRCode\n\nRefer to [./example/qrcode/main.go](./example/qrcode/main.go)\n\n## Testing\n\n```\nAPI_KEY= go test ./...\n```"},{"ref":"P7","kind":"page","title":"novitalabs/python-sdk repository metadata","date":"2026-06-11T04:08:53.991334+00:00","date_source":null,"source_url":"https://github.com/novitalabs/python-sdk","signal_url":null,"signal_json_url":null,"text":"# novitalabs/python-sdk\n\nDescription: Python SDK for Novita AI API (Txt2Img, Img2Img, Txt2Video, Img2Video, Doodle, Remove Background, Replace Object, Reimagine, Merge Faces, ControlNet, VAE, LoRA)\n\nLanguage: Python\n\nLicense: MIT\n\nStars: 27\n\nForks: 8\n\nOpen issues: 0\n\nCreated: 2023-09-27T03:27:34Z\n\nPushed: 2024-11-08T08:50:53Z\n\nDefault branch: main\n\nFork: no\n\nArchived: yes\n\nREADME:\n# Novita AI Python SDK\n\nThis SDK is based on the official [API documentation](https://docs.novita.ai/).\n\n**Join our discord server for help:**\n\n[![](https://dcbadge.vercel.app/api/server/Mqx7nWYzDF)](https://discord.com/invite/Mqx7nWYzDF)\n\n## Installation\n\n```bash\npip install novita-client\n```\n\n## Examples\n\n- [fine tune example](https://colab.research.google.com/drive/1j_ii9TN67nuauvc3PiauwZnC2lT62tGF?usp=sharing)\n- [cleanup](./examples/cleanup.py)\n- [controlnet](./examples/controlnet.py)\n- [img2img](./examples/img2img.py)\n- [img2video](./examples/img2video.py)\n- [inpainting](./examples/inpainting.py)\n- [instantid](./examples/instantid.py)\n- [merge-face](./examples/merge-face.py)\n- [model-search](./examples/model-search.py)\n- [reimagine](./examples/reimagine.py)\n- [remove-background](./examples/remove-background.py)\n- [remove-text](./examples/remove-text.py)\n- [replace-background](./examples/replace-background.py)\n- [txt2img-with-hiresfix](./examples/txt2img-with-hiresfix.py)\n- [txt2img-with-lora](./examples/txt2img-with-lora.py)\n- [txt2img-with-refiner](./examples/txt2img-with-refiner.py)\n- [txt2video](./examples/txt2video.py)\n## Code Examples\n\n### cleanup\n```python\nimport os\n\nfrom novita_client import NovitaClient\nfrom novita_client.utils import base64_to_image\n\nclient = NovitaClient(os.getenv('NOVITA_API_KEY'), os.getenv('NOVITA_API_URI', None))\nres = client.cleanup(\nimage=\"https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo.png\",\nmask=\"https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo_mask.png\"\n)\n\nbase64_to_image(res.image_file).save(\"./cleanup.png\")\n```\n\n### controlnet\n```python\n#!/usr/bin/env python\n# -*- coding: UTF-8 -*-\n\nimport os\n\nfrom novi"},{"ref":"P8","kind":"page","title":"novitalabs/sd-webui-cleaner repository metadata","date":"2026-06-11T04:08:53.953988+00:00","date_source":null,"source_url":"https://github.com/novitalabs/sd-webui-cleaner","signal_url":null,"signal_json_url":null,"text":"# novitalabs/sd-webui-cleaner\n\nDescription: An extension for stable-diffusion-webui to remove any object.\n\nLanguage: JavaScript\n\nLicense: MIT\n\nStars: 344\n\nForks: 26\n\nOpen issues: 5\n\nCreated: 2023-09-27T03:24:18Z\n\nPushed: 2023-10-24T08:06:37Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# Cleaner for Stable Diffusion WebUI\n\n\n\n\n\n\n
\n\n\"Unsplash\"\n\n\nAI image generation API\n
\nNow we have provided the API for remove object. \nCleanup API\n
\n\nThis is a WEBUI extension that provides image erasure functionality. It supports both UI and API simultaneously. Powered by [lama](https://github.com/advimman/lama)\n\n![example1](https://raw.githubusercontent.com/wiki/novitalabs/sd-webui-cleaner/images/example1.png)\n\n
\n\n## Installation\nClone this project in the WEBUI extensions folder\n```\ngit clone https://github.com/novitalabs/sd-webui-cleaner.git\n```\n
\n\n## Get Started\n\nhttps://github.com/novitalabs/sd-webui-cleaner/assets/55743667/3f9f652b-d3b7-4c08-a4c6-0e9fe731c77c\n\n
\n\n### API\n\n```\n//request-----------------------------------\nPOST http://127.0.0.1:7860/cleanup\n\nbody:\n{\n\"input_image\": \"\",\n\"mask\": \"\"\n}\n\n//response-----------------------------------\n{\n\"code\": 0, // 0:success\n\"message\": \"ok\",\n\"image\": \"\"\n}\n```\n\n
\n\n### Used without GPU\nIf you don't have a GPU, please set the cleaner_use_cpu parameter to true through the setting page or api.\n\n
\n\n## Thanks\n- https://github.com/advimman/lama\n- https://github.com/Sanster/lama-cleaner"},{"ref":"P9","kind":"page","title":"novitalabs/javascript-sdk repository metadata","date":"2026-06-11T04:08:53.724925+00:00","date_source":null,"source_url":"https://github.com/novitalabs/javascript-sdk","signal_url":null,"signal_json_url":null,"text":"# novitalabs/javascript-sdk\n\nDescription: JavaScript SDK for Novita AI API (Txt2Img, Img2Img, Txt2Video, Img2Video, Doodle, Remove Background, Replace Object, Reimagine, Merge Faces, ControlNet, VAE, LoRA)\n\nLanguage: TypeScript\n\nStars: 18\n\nForks: 4\n\nOpen issues: 6\n\nCreated: 2023-09-27T15:32:05Z\n\nPushed: 2026-05-27T07:36:41Z\n\nDefault branch: main\n\nFork: no\n\nArchived: yes\n\nREADME:\n\n\n# Novita.ai Javascript SDK\n\nThis SDK is based on the official [novita.ai API reference](https://docs.novita.ai/)\n\n**Join our discord server for help:**\n\n[![](https://dcbadge.vercel.app/api/server/YyPRAzwp7P)](https://discord.gg/YyPRAzwp7P)\n\n## Quick start\n\n1. Sign up on [novita.ai](https://novita.ai) and get an API key. Please follow the instructions at [https://novita.ai/get-started](https://novita.ai/get-started/)\n\n2. Install the [npm package](https://www.npmjs.com/package/novita-sdk) in your project.\n\n```bash\nnpm i novita-sdk\n```\n\n## Version 3.1.0 Update Notes\n\nWe've made significant changes in version 3.0.0. We removed some APIs and will not serve them in the future. The APIs deprecated are:\n\n- adetailer\n- img2mask\n- anymate-anyone\n- create-tile\n- doodle\n- lcm-img2img\n- lcm-txt2img\n- make-photo\n- mix-pose\n- relight\n- remove-watermark\n- replace-sky\n- replace-object\n- upscale\n- img2prompt\n- img2video-motion\n- outpainting\n- reimagine\n- restore-face\n\n## Usage\n\n```javascript\nimport { NovitaSDK } from \"novita-sdk\";\n\nconst novitaClient = new NovitaSDK(\"your api key\");\n\nconst params = {\nrequest: {\nmodel_name: \"majicmixRealistic_v7_134792.safetensors\",\nprompt: \"1girl,sweater,white background\",\nnegative_prompt: \"(worst quality:2),(low quality:2),(normal quality:2),lowres,watermark,\",\nwidth: 512,\nheight: 768,\nsampler_name: \"Euler a\",\nguidance_scale: 7,\nsteps: 20,\nimage_num: 1,\nseed: -1,\n},\n};\nnovitaClient\n.txt2Img(params)\n.then((res) => {\nif (res && res.task_id) {\nconst timer = setInterval(() => {\nnovitaClient\n.progress({\ntask_id: res.task_id,\n})\n.then((progressRes) => {\nif (progressRes.task.status === TaskStatus.SUCCEED) {\nconsole.log(\"finished!\", progressRes.images);\nclearInterval(timer);\nonFinish(progressRes.images);\n}\nif (progressRes.task.status === TaskStatus.FAILED) {\nco"},{"ref":"P10","kind":"page","title":"novitalabs/AnimateAnyone repository metadata","date":"2026-06-11T04:08:53.491211+00:00","date_source":null,"source_url":"https://github.com/novitalabs/AnimateAnyone","signal_url":null,"signal_json_url":null,"text":"# novitalabs/AnimateAnyone\n\nDescription: Unofficial Implementation of Animate Anyone by Novita AI\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 779\n\nForks: 69\n\nOpen issues: 6\n\nCreated: 2024-05-30T08:09:27Z\n\nPushed: 2026-05-27T06:27:09Z\n\nDefault branch: main\n\nFork: no\n\nArchived: yes\n\nREADME:\n# Animate Anyone\n\n[![Novita AI](https://github.com/novitalabs/AnimateAnyone/assets/4327933/2a6ef880-e5c3-437e-adc5-2ae8601ac4f8)](https://novita.ai)\n\n## Overview\n\nThis repository currently provides the unofficial pre-trained weights and inference code of [Animate Anyone](https://humanaigc.github.io/animate-anyone). It is inspired by the implementation of the [MooreThreads/Moore-AnimateAnyone](https://github.com/MooreThreads/Moore-AnimateAnyone) repository and we made some adjustments to the training process and datasets.\n\n## Samples\n\n\n\n\n\n\n
\n\n## Quickstart\n\n### Build Environtment\n\nWe Recommend a python version `>=3.10` and cuda version `=11.7`. Then build environment as follows:\n\n```shell\n# [Optional] Create a virtual env\npython -m venv .venv\nsource .venv/bin/activate\n# Install with pip:\npip install -r requirements.txt\n```\n\n### Download weights\n\n**Automatically downloading**: You can run the following command to download weights automatically:\n\n```shell\npython tools/download_weights.py\n```\n\nWeights will be placed under the `./pretrained_weights` direcotry. The whole downloading process may take a long time.\n\n### Inference\n\nHere is the cli command for running inference scripts:\n\n```shell\npython -m scripts"},{"ref":"P11","kind":"page","title":"novitalabs/Novita-CollabHub repository metadata","date":"2026-06-11T04:08:53.488714+00:00","date_source":null,"source_url":"https://github.com/novitalabs/Novita-CollabHub","signal_url":null,"signal_json_url":null,"text":"# novitalabs/Novita-CollabHub\n\nStars: 6\n\nForks: 2\n\nOpen issues: 0\n\nCreated: 2024-08-29T08:00:00Z\n\nPushed: 2026-05-08T10:05:37Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n
\n\"Novita\n\n[𝕏 Follow me on X](https://x.com/novita_labs?utm_source=github_collabhub&utm_medium=readme&utm_campaign=collab) • [🤗 Hugging Face](https://huggingface.co/novita?utm_source=github_collabhub&utm_medium=readme&utm_campaign=collab) • [💻 Docs](https://novita.ai/docs/guides/introduction?utm_source=github_collabhub&utm_medium=readme&utm_campaign=collab)\n\n[Novita AI](https://novita.ai?utm_source=github_collabhub&utm_medium=readme&utm_campaign=collab) is an AI cloud platform that helps developers easily deploy AI models through a simple API, backed by affordable and reliable GPU cloud infrastructure.\n
\n\n## **TOP-LLM Integration Repo**\n\nIntegrate the Novita API into popular software and platforms. Access [Novita](https://novita.ai/settings/key-management?utm_source=github_collabhub&utm_medium=readme&utm_campaign=collab) to get an API key.\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
RepoDescriptionGuide
Hugging Face is a library that provides pre-trained language models for NLP tasks. Novita is one of the inference providers of Hugging Face.Novita AI & Hugging Face Integration Guide
LangChain is a framework for developing applications powered by large language models (LLMs).Novita AI & LangChain Integration Gui"},{"ref":"P12","kind":"page","title":"novitalabs/dify-plugin-novita repository metadata","date":"2026-06-11T04:08:53.217279+00:00","date_source":null,"source_url":"https://github.com/novitalabs/dify-plugin-novita","signal_url":null,"signal_json_url":null,"text":"# novitalabs/dify-plugin-novita\n\nLanguage: Python\n\nStars: 0\n\nForks: 0\n\nOpen issues: 0\n\nCreated: 2025-08-13T06:40:15Z\n\nPushed: 2026-06-11T03:02:24Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n## Overview\n\n[Novita AI](https://novita.ai/) is an AI cloud platform that helps developers easily deploy AI models through a simple API, backed by affordable and reliable GPU cloud infrastructure.\n\nNovita AI supports various models from DeepSeek, Llama, Qwen, GLM, etc. Check all models [here](https://novita.ai/models/llm).\n\n# Configuration\n\n1. [Create a Novita AI account.](https://novita.ai/user/login)\n2. Create and save your API key [here](https://novita.ai/settings/key-management).\n- Click **Add New Key** to generate a new API Key. Note that the key **will only be displayed once upon generation** — ensure to save it in a safe place.\n![](_assets/novita-02.png)\n\n3. Install the plugin in Dify, access the [Settings] page to enter the API key you just created. \n![](_assets/novita-01.png)\n\n4. Now you can use the Novita AI models in Dify.\n\nSource code of this plugin: [https://github.com/novitalabs/dify-plugin-novita](https://github.com/novitalabs/dify-plugin-novita)"},{"ref":"P13","kind":"page","title":"novitalabs/novita-mcp-server repository metadata","date":"2026-06-11T04:08:53.213138+00:00","date_source":null,"source_url":"https://github.com/novitalabs/novita-mcp-server","signal_url":null,"signal_json_url":null,"text":"# novitalabs/novita-mcp-server\n\nDescription: The Model Context Protocol (MCP) server that provides seamless interaction with Novita AI platform resources\n\nLanguage: JavaScript\n\nLicense: MIT\n\nStars: 11\n\nForks: 9\n\nOpen issues: 3\n\nCreated: 2025-05-06T11:58:54Z\n\nPushed: 2025-05-12T09:53:05Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# Novita MCP Server\n[![smithery badge](https://smithery.ai/badge/@novitalabs/novita-mcp-server)](https://smithery.ai/server/@novitalabs/novita-mcp-server)\n\n`novita-mcp-server` is a Model Context Protocol (MCP) server that provides seamless interaction with Novita AI platform resources. We recommend accessing this server through [Claude Desktop](https://claude.ai/download), [Cursor](https://www.cursor.com/), or any other compatible MCP client.\n\n\n\"Novita\n\n\n## Features\n\n> ⚠️ **Beta Notice**: `novita-mcp-server` is currently in beta and only supports GPU instance management. Additional resource types will be supported in future releases.\n\nCurrently, `novita-mcp-server` enables management the resources of [GPU instances product](https://novita.ai/gpus-console). \n\nSupported operations are as follows:\n- Cluster(/Region): List;\n- Product: List;\n- GPU Instance: List, Get, Create, Start, Stop, Delete, Restart;\n- Template: List, Get, Create, Delete;\n- Container Registry Auth: List, Create, Delete;\n- Network Storage: List, Create, Update, Delete;\n\n## Installation\n\nYou can install the package using npm, or Smithery:\n\n**Using npm**\n\n```bash\nnpm install -g @novitalabs/novita-mcp-server\n```\n\n**Using Smithery**\n\nVisit the [https://smithery.ai/server/@novitalabs/novita-mcp-server](https://smithery.ai/server/@novitalabs/novita-mcp-server) and follow the \"Install\" instructions to install the server.\n\n## Configuration to use novita-mcp-server\n\nFirst, you need to get your Novita API key from the [Novita AI Key Management](https://novita.ai/settings/key-management).\n\nAnd next, you can use the following configuration for both Claude Desktop and Cursor:\n"},{"ref":"P14","kind":"page","title":"novitalabs/chatbynovita repository metadata","date":"2026-06-11T04:08:52.152565+00:00","date_source":null,"source_url":"https://github.com/novitalabs/chatbynovita","signal_url":null,"signal_json_url":null,"text":"# novitalabs/chatbynovita\n\nStars: 0\n\nForks: 1\n\nOpen issues: 0\n\nCreated: 2025-08-30T18:22:32Z\n\nPushed: 2025-08-19T10:09:57Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME: none published or not readable through the GitHub API."},{"ref":"P15","kind":"page","title":"novitalabs/autotuner repository metadata","date":"2026-06-11T04:08:51.790688+00:00","date_source":null,"source_url":"https://github.com/novitalabs/autotuner","signal_url":null,"signal_json_url":null,"text":"# novitalabs/autotuner\n\nDescription: Optimize the performance of LLM inference engines by automatically tuning parameters for a specific model.\n\nLanguage: Python\n\nLicense: MIT\n\nStars: 11\n\nForks: 3\n\nOpen issues: 4\n\nCreated: 2025-10-21T11:17:53Z\n\nPushed: 2026-06-10T20:58:22Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# LLM Autotuner (for inference)\n\n

\n\"Autotuner\n

\n\nAutomated parameter tuning for LLM inference engines (SGLang, vLLM) for best performance, while respecting SLOs and hardware constraints.\n\n## Why Autotuner?\n\n**Quantization and parameter tuning can unlock 60%+ performance gains.** LLM inference engines like SGLang and vLLM ship with conservative defaults that work everywhere but are optimized for nowhere.\n\n### Performance Impact: Real-World Data\n\n

\n\"Throughput\n

\n\n

\n\"Latency\n

\n\nTesting on NVIDIA RTX 4090 (24GB) with typical production workloads (mixed prefill/decode).\n\n**See detailed benchmarks:** [Baseline Benchmarks](docs/qwen-benchmarks.md)\n\n| What You Get | Manual Tuning | Autotuner |\n|--------------|---------------|-----------|\n| **Time to optimal config** | Hours to Days | **Minutes** |\n| **Parameter combinations tested** | ~10 (limited by patience) | **50-100+** (automated) |\n| **Performance gain** | Unknown (untested) | **60%+ throughput** (quantization + tuning) |\n| **Reproducibility** | Low (manual errors) | **High** (versioned configs) |\n| **Cross-hardware portability** | Manual rework | **Re-run task** (one command) |\n\n## How to Use\n\n### CLI Mode\n

\n\"CLI\n

\n\n### Web UI Mode\n

\n\"Web\n

\n\n### Agent Mode\n

\n\"Agent\n

\n\n## Core Concepts\n

\n\"Cor"},{"ref":"P16","kind":"page","title":"novitalabs/javascript-sdk\n\n

KV cache on the wings of Pegasus.

\n\n[![CI](https://github.com/novitalabs/pegaflow/actions/workflows/ci.yml/badge.svg)](https://github.com/novitalabs/pegaflow/actions/workflows/ci.yml)\n[![PyPI](https://img.shields.io/pypi/v/pegaflow-llm)](https://pypi.org/project/pegaflow-llm/)\n[![License](https://img.shields.io/badge/license-Apache--2.0-blue)](LICENSE)\n\n\n**PegaFlow is a high-performance KV cache storage engine for LLM inference.** Offload KV cache from GPU to host memory or SSD, and share it across nodes via RDMA.\n\n- **Decoupled from inference lifecycle** — runs as an independent sidecar; KV cache survives engine restarts, scales independently, and is shared across instances\n- **Topology-aware, PCIe-saturating transfers** — NUMA-aware pinned memory + layer-wise DMA to maximize hardware bandwidth\n- **GIL-free Rust core** — zero Python overhead on the hot path; your inference engine keeps its threads\n- **Production-ready observability** — built-in Prometheus metrics and OTLP export, not an afterthought\n- **Pluggable** — works with vLLM as a drop-in KV connector\n\n## News\n\n- **2026-05-18** — [vLLM x Novita AI: PegaFlow for Production-Grade External KV Cache](https://vllm.ai/blog/2026-05-18-pegaflow), a joint blog post with the vLLM team.\n\n## Architecture\n\n
\n\"PegaFlow\n
\n\n## Framework Integration\n\n| Framework | Status | Link |\n|-----------|--------|------|\n| vLLM | ✅ Ready | [Quick Start](#3-launch-your-inference-engine) |\n\n## Quick Start\n\n### 1. Install\n\n```bash\nuv pip install pegaflow-llm # CUDA 12\nuv pip install pegaflow-llm-cu13 # CUDA 13\n```\n\n### 2. Start PegaFlow Server\n\n```bash\npegaflow-server\n```\n\n### 3. La"},{"ref":"P26","kind":"page","title":"novitalabs/rft-tinker repository metadata","date":"2026-06-11T03:02:48.009234+00:00","date_source":null,"source_url":"https://github.com/novitalabs/rft-tinker","signal_url":null,"signal_json_url":null,"text":"# novitalabs/rft-tinker\n\nDescription: RL training demo with tinker + sandbox\n\nLanguage: Python\n\nStars: 0\n\nForks: 0\n\nOpen issues: 0\n\nCreated: 2026-01-19T08:22:11Z\n\nPushed: 2026-01-26T03:06:14Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# RFT-Tinker: R2E-Gym Training with Tinker API + Agent Sandbox\n\n## Overview\n\nExperimental setup for training code generation models on R2E-Gym dataset using:\n- **Tinker API** for RL model training\n- **Agent Sandbox** for safe code execution\n- **R2E-Gym Dataset** (4.5K real-world GitHub issues)\n\nReproducing DeepSWE experiments (42.2% Pass@1 on SWE-Bench-Verified).\n\n## Quick Start\n\n### 1. Clone Repository\n\n```bash\ngit clone https://github.com/novitalabs/rft-tinker.git\ncd rft-tinker\n```\n\n### 2. Install Dependencies\n\n```bash\npython3 -m venv venv\nsource venv/bin/activate\npip install datasets huggingface-hub novita-sandbox tinker torch transformers\n```\n\n### 3. Configure API Keys\n\nCopy the example environment file:\n\n```bash\ncp .env.example .env.local\n```\n\nEdit `.env.local` with your API keys:\n\n```bash\n# Agent Sandbox API Key (get from https://novita.ai)\nNOVITA_API_KEY=your_novita_api_key_here\n\n# Tinker API Token (get from Tinker platform)\nTINKER_API_TOKEN=your_tinker_api_token_here\n\n# Template IDs\nNOVITA_TEMPLATE_BASE=vn9xnp3cm92x6rmqlgwc\n```\n\n**Warning: Never commit `.env.local` with real credentials!**\n\n### 4. Run Tests\n\nTest Agent Sandbox connectivity:\n```bash\npython -m tests.integration.test_novita_basic\n```\n\nTest R2E-Gym workflow:\n```bash\npython -m tests.integration.test_r2e_gym_workflow\n```\n\n### 5. Prepare Dataset\n\nDownload R2E-Gym sample (50 instances):\n```bash\npython scripts/prepare_data/prepare_r2e_sample.py\n```\n\nTest dataset loading:\n```bash\npython -m tests.unit.test_dataset_loading\n```\n\n## Project Structure\n\n```\nrft-tinker/\n├── src/ # Core source code\n│ ├── datasets/ # Dataset utilities and repo mapping\n│ ├── environments/ # Sandbox environment wrappers\n│ ├── rollout/ # Multi-turn rollout pipeline\n│ └── utils/ # Utility functions\n├── tests/ # All test files\n│ ├── integration/ # Integration tests\n│ ├── rollout/ # Rollout pipeline tests\n│ └── unit/ # Unit tests\n├── scripts/ # Utility scripts\n├── templates/ # Agent San"},{"ref":"P27","kind":"page","title":"novitalabs/pegaflow v0.22.5","date":"2026-06-11T02:59:38.098498+00:00","date_source":null,"source_url":"https://github.com/novitalabs/pegaflow/releases/tag/v0.22.5","signal_url":null,"signal_json_url":null,"text":"# v0.22.5\n\nRepository: novitalabs/pegaflow\n\nTag: v0.22.5\n\nPublished: 2026-06-02T09:57:38Z\n\nPrerelease: no\n\nRelease notes: none published."},{"ref":"P28","kind":"page","title":"novitalabs/dify-plugins repository metadata","date":"2026-06-11T02:53:57.234072+00:00","date_source":null,"source_url":"https://github.com/novitalabs/dify-plugins","signal_url":null,"signal_json_url":null,"text":"# novitalabs/dify-plugins\n\nDescription: All Dify Plugins listed in Dify Marketplace, plus illustrated plugin examples.\n\nStars: 0\n\nForks: 0\n\nOpen issues: 0\n\nCreated: 2025-08-27T07:32:20Z\n\nPushed: 2025-08-27T07:49:10Z\n\nDefault branch: main\n\nFork: yes\n\nParent repository: langgenius/dify-plugins\n\nArchived: no\n\nREADME:\n![Dify Logo](.assets/bar.jpeg)\n\n

\nDify Cloud ·\nSelf-hosting ·\nDocumentation ·\nEnterprise inquiry\n

\n\n

\n\n\"Static\n\n\"Static\n\n\n \n\n\n\n\n\n\n\"Docker\n\n