Google's source-grounded research notebook, now presented as Gemini Notebook at notebook.google. Analyzes your files and turns them into audio, video, and more.

Anthropic's previous Opus snapshot: 200k context, $5/$25 per 1M tokens. Later billed lines include Opus 4.6 and Opus 5.

Meta ANN library. A library, not a hosted vector DB or marketing ally.

MiroFish is an open-source swarm intelligence engine that spins up agent personas around a news event to surface second-order effects, not a single forecast.

Citation skill that searches Google Scholar and PubMed, turns DOIs, PMIDs, and arXiv IDs into BibTeX, and validates bibliographies before submission.

K-Dense agent skill that turns observations into rival hypotheses, discriminating predictions, and preregistration-ready analysis plans.

Research poster skill for beamerposter, tikzposter, or baposter, with AI-generated visuals under hard readability limits and an overflow check before print.

Systematic literature review skill: multi-database search via parallel-cli, PRISMA-style screening, DOI citation checks, and Markdown or PDF output.

Generate comprehensive 50+ page market research reports in consulting firm style - extensive visuals, deep research-lookup integration, multi-framework strategic analysis.

Confidentiality-first peer review skill: authorization gate, claim-evidence and statistics checks, and local CLIs that draft and lint a review.

Grant-writing skill for NSF, NIH, DOE, DARPA, and Taiwan NSTC proposals: agency review criteria, specific aims, broader impacts, budgets, and a phased timeline.

Manuscript research skill that uses Parallel Search and Extract to build a packet of up to 60 verified references, with Perplexity as an opt-in fallback.

Apply ScholarEval framework to systematically evaluate scholarly work - comprehensive analysis across research quality dimensions with structured scoring and feedback.

Diagram skill that generates scientific figures with Nano Banana 2 and regenerates only when a Gemini 3.6 Flash review scores below your document type's bar.

Research talk skill that plans a deck, generates each slide as an AI image through OpenRouter, and combines them into a PDF, with a PowerPoint path.

Evidence-bound manuscript skill: every claim maps to verified sources, with reporting-guideline selection and offline consistency and citation checks.

Verification-first skill for journal, conference, poster, and grant formatting, with bundled LaTeX scaffolds and a PDF page and font checker.

Praxist is a Fair Source research runtime that runs parallel peers, durable evidence, and Codex or Claude Code skills on a measurable project.

Apple's 9B vision-language model reads documents as compressed images, then expands only the pages that matter, holding accuracy to 4.3x compression.

BAAI's 27B open-weight agent model for long-horizon self-improvement: reflect, measure, and revise across coding, ML engineering, and deep research with a 262K context.

Bilibili's open multilingual translation family: 150-language text, speech subtitles, controlled dubbing, and long-document translation built on Qwen3.5.

From-scratch English WWII language models at 291M and 148M parameters, with Apache-2.0 weights and CPU demos; not a RAG assistant.

AI research workspace for paper reading, academic writing, data analysis, and interview transcription, with free and paid plans.

Hugging Face's paper-publisher skill indexes arXiv papers on the Hub, links them to models and datasets, claims authorship and drafts articles.