Support Jamba from AI21 Studio

.github/workflows/test-ai21.yml

@@ -0,0 +1,26 @@
+name: Test on AI21
+
+on:
+  push:
+    branches: [main]
+  workflow_dispatch:
+
+jobs:
+  test:
+    runs-on: ubuntu-22.04
+    timeout-minutes: 5
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version: '18'
+
+      - run: echo 'Which planet in our solar system is the largest?' | ./ask-llm.js | tee output.txt | grep -i jupiter
+        timeout-minutes: 3
+        env:
+          LLM_API_BASE_URL: 'https://api.ai21.com/studio/v1'
+          LLM_API_KEY: ${{ secrets.AI21_API_KEY }}
+          LLM_CHAT_MODEL: 'jamba-1.5-mini'
+
+      - run: cat output.txt

README.md

@@ -62,6 +62,7 @@ export LLM_API_BASE_URL=http://localhost:3928/v1
 
 ## Using Managed LLM Services
 
+[![Test on AI21](https://github.com/ariya/ask-llm/actions/workflows/test-ai21.yml/badge.svg)](https://github.com/ariya/ask-llm/actions/workflows/test-ai21.yml)
 [![Test on DeepInfra](https://github.com/ariya/ask-llm/actions/workflows/test-deepinfra.yml/badge.svg)](https://github.com/ariya/ask-llm/actions/workflows/test-deepinfra.yml)
 [![Test on Fireworks](https://github.com/ariya/ask-llm/actions/workflows/test-fireworks.yml/badge.svg)](https://github.com/ariya/ask-llm/actions/workflows/test-fireworks.yml)
 [![Test on Groq](https://github.com/ariya/ask-llm/actions/workflows/test-groq.yml/badge.svg)](https://github.com/ariya/ask-llm/actions/workflows/test-groq.yml)
@@ -73,10 +74,16 @@ export LLM_API_BASE_URL=http://localhost:3928/v1
 [![Test on OpenRouter](https://github.com/ariya/ask-llm/actions/workflows/test-openrouter.yml/badge.svg)](https://github.com/ariya/ask-llm/actions/workflows/test-openrouter.yml)
 [![Test on Together](https://github.com/ariya/ask-llm/actions/workflows/test-together.yml/badge.svg)](https://github.com/ariya/ask-llm/actions/workflows/test-together.yml)
 
-Supported LLM services include [Deep Infra](https://deepinfra.com), [Fireworks](https://fireworks.ai), [Groq](https://groq.com), [Hyperbolic](https://www.hyperbolic.xyz), [Lepton](https://lepton.ai), [Novita](https://novita.ai), [Octo](https://octo.ai), [OpenAI](https://platform.openai.com), [OpenRouter](https://openrouter.ai), and [Together](https://www.together.ai).
+Supported LLM services include [AI21](https://studio.ai21.com), [Deep Infra](https://deepinfra.com), [Fireworks](https://fireworks.ai), [Groq](https://groq.com), [Hyperbolic](https://www.hyperbolic.xyz), [Lepton](https://lepton.ai), [Novita](https://novita.ai), [Octo](https://octo.ai), [OpenAI](https://platform.openai.com), [OpenRouter](https://openrouter.ai), and [Together](https://www.together.ai).
 
 For configuration specifics, refer to the relevant section. The examples use Llama-3.1 8B (or GPT-4o Mini for OpenAI), but any LLM with at least 7B parameters should work just as well, such as Mistral 7B, Qwen-2 7B, or Gemma-2 9B.
 
+* [AI21](https://studio.ai21.com)
+```bash
+export LLM_API_BASE_URL=https://api.ai21.com/studio/v1
+export LLM_API_KEY="yourownapikey"
+export LLM_CHAT_MODEL=jamba-1.5-mini
+```
 
 * [Deep Infra](https://deepinfra.com)
 ```bash