Source: https://seerepo.com/compare/comet-ml/opik/vs/keras-team/keras
Markdown source: https://seerepo.com/compare/comet-ml/opik/vs/keras-team/keras.md
Description: opik and keras share 4 technologies: Rich, torch, scikit-learn, and pillow. Both are written in Python.

# opik vs keras

opik and keras share 4 technologies: Rich, torch, scikit-learn, and pillow. Both are written in Python.

- [comet-ml/opik](https://seerepo.com/r/comet-ml/opik): Debug, evaluate, and monitor your LLM applications, RAG systems, and agentic workflows with comprehensive tracing, automated evaluations, and production-ready dashboards.
- [keras-team/keras](https://seerepo.com/r/keras-team/keras): Deep Learning for humans

## What they share and what sets them apart

- In common: Rich, torch, scikit-learn, pillow
- Only in opik: FastAPI, JAX-RS, Flask, werkzeug, flask-cors, redisson, s3, Redis, MySQL, OpenAI, Anthropic, Sentry, AWS S3, Hugging Face, AIOHTTP, anyio, Vite, React, tanstack, PostHog, itsdangerous, accelerate, spaCy, tiktoken, tokenizers, litellm, LangChain, transformers, datasets, peft, markdown-it-py, markupsafe, jinja2, annotated-types, referencing, Pydantic, jsonschema, pydantic-core, Gunicorn, OpenTelemetry, opentelemetry-instrumentation-flask, opentelemetry-instrumentation-logging, opentelemetry-instrumentation-requests, opentelemetry-instrumentation-system-metrics, botocore, docker, class-variance-authority, Tailwind CSS, tailwind-merge, React Hook Form, Radix UI, @uiw/react-codemirror, CodeMirror, recharts, dnd-kit, react-hotkeys-hook, react-textarea-autosize, react18-json-view, react-day-picker, cmdk, react-intersection-observer, react-player, @lezer/highlight, Mermaid, re-resizable, react-grid-layout, react-resizable-panels, Vega, vega-lite, Lucide, mustache, react-markdown, remark-breaks, remark-gfm, ansi-to-html, react-pdf, rehype-raw, rehype-sanitize, unist-util-visit, yaml, Zod, @hookform/resolvers, ora, ai, @ai-sdk/google, Nunjucks, @clack/prompts, zod-to-json-schema, @opentelemetry/sdk-trace-base
- Only in keras: h5py, TensorFlow, ml-dtypes, ai-edge-litert, onnxruntime

## Side by side

| | opik | keras |
|---|---|---|
| Size | 1,680,031 lines written, 7,878 code files | 324,787 lines written, 624 code files |
| Who writes it | 19.7% of its commits signed by AI. With Claude Code, GitHub Copilot, and Cursor | 0.5% of its commits signed by AI. With Gemini CLI, Jules, Claude Code, and Cursor |
| Activity | Last push November 4, 2026, 7,898 commits, 23,166 stars | Last push November 5, 2026, 11,732 commits, 64,419 stars |
| Tests | 2,703 test files. 20.9% of its code files have a test that imports them | 318 test files. 34.8% of its code files have a test that imports them |

- Size: Lines written by people and agents, without generated ones; code files, without tests.
- Who writes it: A commit counts as AI's when it carries its signature: a Co-Authored-By line, its author name or the merge of its branch.
- Activity: From GitHub and its history, on the day SeeRepo read it.
- Tests: The files that test the code, and the share of code files a test imports.

## Frequently asked questions

### Which is bigger, opik or keras?

opik, with 1,680,031 lines written; keras has 324,787.

The proof: [comet-ml/opik's code at commit df61dd2](https://github.com/comet-ml/opik/tree/df61dd21c66e5ed901dafba326bf29cb4b4fbcfb)

### Which has more GitHub stars, opik or keras?

keras, with 64,419 stars; opik has 23,166.

The proof: [keras-team/keras on GitHub](https://github.com/keras-team/keras)

### Which has more tests, opik or keras?

opik, with 2,703 test files; keras has 318.

The proof: [comet-ml/opik's code at commit df61dd2](https://github.com/comet-ml/opik/tree/df61dd21c66e5ed901dafba326bf29cb4b4fbcfb)

### Which was updated more recently, opik or keras?

keras: its last push was on November 5, 2026, and opik's on November 4, 2026.

The proof: [keras-team/keras's commits on GitHub](https://github.com/keras-team/keras/commits)

### Which has more commits signed by AI, opik or keras?

opik: an AI signed 19.7% of its commits, with Claude Code, GitHub Copilot, and Cursor; in keras, 0.5%.

The proof: [the commits each tool signed in comet-ml/opik](https://seerepo.com/r/comet-ml/opik/ai)

## Sources

- comet-ml/opik: Read on November 4, 2026, at commit df61dd2 (https://github.com/comet-ml/opik/tree/df61dd21c66e5ed901dafba326bf29cb4b4fbcfb)
- keras-team/keras: Read on November 5, 2026, at commit 104de14 (https://github.com/keras-team/keras/tree/104de145739fc5e6c64504edb3ab3736a645ff1b)

## What is your repo built with?

Any public repo, with no account: https://seerepo.com/r/owner/repo/built, with the owner and name it has on GitHub.