# Berri AI — developer ecosystem report

> The fastest way to take your LLM app to production

- GitHub org: `BerriAI`
- Category: ai
- Overall score: 49.0 / 100
- Overall rank: #240
- Last refreshed: 2026-08-20

## Pillar scores

| Pillar | Score / 100 | Rank | Weight in overall |
| --- | --- | --- | --- |
| Code & Polish | 43.0 | #645 | ~22% |
| Education | 70.0 | #351 | ~18% |
| AI readiness | 60.0 | #88 | part of Reach |
| Builder experience | 0.0 | #1 | ~12% |
| Community | 49.0 | #177 | ~16% |
| Reach | 34.0 | #670 | ~18% |
| Momentum | 50.0 | #222 | ~14% |

## Links

- GitHub: https://github.com/BerriAI
- Website: https://docs.litellm.ai
- Docs: https://docs.litellm.ai/docs

## Public GitHub totals

| Metric | Value |
| --- | --- |
| Repos analyzed | 12 |
| Stars | 57,571 |
| Forks | 11,168 |
| Open issues | 5,257 |
| Contributors | — |

## Top repositories

| Repo | Stars | Language | Description |
| --- | --- | --- | --- |
| [litellm](https://github.com/BerriAI/litellm) | 56,811 | Python | The fastest, litest AI Gateway. Rust core with Python SDK. Call 100+ LLM APIs in OpenAI (or native) format with cost tra |
| [liteLLM-proxy](https://github.com/BerriAI/liteLLM-proxy) | 255 | Python | — |
| [clerkie-cli](https://github.com/BerriAI/clerkie-cli) | 102 | Python | Clerkie-cli is a terminal tool that uses an LLM to help debug and troubleshoot coding errors in any programming language |
| [litellm-pgvector](https://github.com/BerriAI/litellm-pgvector) | 84 | Python | — |
| [litellm-skills](https://github.com/BerriAI/litellm-skills) | 77 | Python | Agent Skills for managing live LiteLLM proxy deployments — users, teams, keys, orgs, models, MCP servers, agents |
| [terraform-provider-litellm](https://github.com/BerriAI/terraform-provider-litellm) | 38 | Go | litellm terraform provider |
| [self-improving-agent](https://github.com/BerriAI/self-improving-agent) | 24 | TypeScript | Drop-in self-improvement loop for AI agents — two tools + a skill. Agent proposes a diff, human approves, draft PR opens |
| [litellm-docs](https://github.com/BerriAI/litellm-docs) | 19 | JavaScript | The official LiteLLM docs repository |
| [ai-gateway-bench](https://github.com/BerriAI/ai-gateway-bench) | 17 | Python | Benchmarks for evaluating ai-gateways |
| [litellm-guardrails](https://github.com/BerriAI/litellm-guardrails) | 15 | Svelte | Registry of public custom code guardrails for the litellm proxy server |
| [example_openai_endpoint](https://github.com/BerriAI/example_openai_endpoint) | 13 | Python | An example OpenAI /chat/completions endpoint  |
| [shin-watcher-oss](https://github.com/BerriAI/shin-watcher-oss) | 10 | TypeScript | shin-builder-oss |

## Methodology

Scores are computed from public signals only (GitHub, package registries, docs sites, community platforms). Pillar definitions: https://www.smoower.com/docs.md

---

HTML version: https://www.smoower.com/berriai
Machine index: https://www.smoower.com/llms.txt · MCP server: https://www.smoower.com/mcp · REST API: https://www.smoower.com/api/v1