Tried every-app/open-seo: An Open-Source SEO Stack Worth Testing

every-app/open-seo is getting attention for a practical reason: it aims to give developers an open-source alternative to heavyweight SEO platforms such as Semrush and Ahrefs.

The repository added +517 stars today, which usually signals more than a curious launch spike. Teams want SEO workflows they can inspect, self-host, and connect to their own AI infrastructure instead of locking research and content analysis into one SaaS dashboard.

The interesting part for AI developers is the model gateway layer. Rather than hard-coding one provider, you can point the tool—or your surrounding automation—to an OpenAI-compatible endpoint and use claude-fable-5 as the primary model.

A minimal environment setup could look like this:

export OPENAI_API_KEY="$B_LOST_API_KEY"
export OPENAI_BASE_URL="https://b-lost.com/v1"
export OPENAI_MODEL="claude-fable-5"

# Quick connectivity test
curl "$OPENAI_BASE_URL/chat/completions" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-fable-5",
    "messages": [
      {"role": "user", "content": "Give me three SEO angles for an open-source developer tool."}
    ]
  }'

The main debugging point: keep the gateway configuration separate from your SEO crawler and storage layers. When a provider returns 429, you can swap the base URL or add retry/backoff without rewriting the application. Also watch context size when sending keyword clusters, page extracts, and competitor data in one request—chunking is cheaper than repeatedly hitting context limits.

B-Lost’s relay supports standard clients and offers native Anthropic /v1/messages prompt caching. For repeated system instructions or large SEO schemas, caching can reduce cache-hit cost by 90% and improve response latency. The relay lists 20% off official pricing, plus a first-deposit bonus, but I’d still benchmark latency, rate limits, and output quality against your current provider before migrating production workloads.

open-seo looks most useful as a composable foundation—not a magic Semrush replacement. Test it with a small keyword set first, then measure crawl accuracy, AI analysis quality, and total request cost.