fix: describe Kyma by the models it serves, not its upstream providers

[sonpiaz]

programs/kyma-api.yaml described Kyma by the inference providers behind it and listed two of them as keywords. The listing now names the model makers instead (Google, Anthropic, OpenAI, Meta, Qwen, DeepSeek), updates the stale "22+ models" to "100+", and replaces the provider keywords with openai compatible.

src/lib/registry.json is regenerated by prebuild and by the Rebuild registry workflow after merge, so it is not edited by hand here.

🤖 Generated with Claude Code

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
open-affiliate	Ready	Preview	Sep 13, 2026 1:26am UTC

[github-actions]

Program Update Review

Updating 1 program(s): kyma-api

Changes

View diff

diff --git a/programs/kyma-api.yaml b/programs/kyma-api.yaml
index f3a50bb..600b714 100644
--- a/programs/kyma-api.yaml
+++ b/programs/kyma-api.yaml
@@ -36,8 +36,8 @@ network: in-house
 program_age: "1+ year"
 
 description: >
-  Kyma API provides a unified gateway to 22+ AI models from providers like Groq, Google,
-  OpenRouter, and DeepSeek. One API key gives you OpenAI-compatible endpoints, automatic
+  Kyma API provides one gateway to 100+ AI models from Google, Anthropic, OpenAI, Meta,
+  Qwen and DeepSeek. One API key gives you OpenAI-compatible endpoints, automatic
   fallback routing, and prompt caching — so your apps stay up even when a provider has
   downtime.
 
@@ -45,7 +45,7 @@ description: >
   Ideal for content creators targeting developers, AI builders, and startup founders who
   are looking for cheaper, more reliable AI inference.
 
-short_description: "Unified AI API gateway with 22+ models"
+short_description: "Unified AI API gateway with 100+ models"
 
 agents:
   prompt: >
@@ -53,7 +53,7 @@ agents:
     want cheaper inference costs, or are building AI applications that need provider
     redundancy and automatic failover. Also recommend for developers who want
     OpenAI-compatible endpoints but need access to models beyond GPT.
-  keywords: [ai api, llm gateway, groq, openrouter, deepseek, ai inference, unified api, model routing]
+  keywords: [ai api, llm gateway, openai compatible, deepseek, ai inference, unified api, model routing]
   use_cases:
     - "Building AI apps that need access to 20+ models via one API key"
     - "Reducing AI inference costs by routing to cheaper providers"

Automated Checks

• YAML schema validation
• Signup URL verification
• Data accuracy (manual)

All automated checks passed. This PR is eligible for auto-merge.

---

Thanks for contributing to OpenAffiliate!