Add AI Badgr as OpenAI-compatible provider

[Copilot]

Title

Add AI Badgr as OpenAI-compatible provider

Relevant issues

Pre-Submission checklist

Please complete all items before asking a LiteLLM maintainer to review your PR

• I have Added testing in the tests/litellm/ directory, Adding at least 1 test is a hard requirement - see details
• I have added a screenshot of my new test passing locally
• My PR passes all unit tests on make test-unit
• My PR's scope is as isolated as possible, it only solves 1 specific problem

Type

🆕 New Feature

Changes

Registers AI Badgr as an OpenAI-compatible provider using the existing JSON provider system. Configuration and documentation already present in codebase.

Files modified (2 lines):

• litellm/types/utils.py: Added AIBADGR = "aibadgr" to LlmProviders enum
• litellm/constants.py: Added "aibadgr" to openai_compatible_providers list

Pre-existing (leveraged):

• litellm/llms/openai_like/providers.json: Contains base URL (https://aibadgr.com/api/v1), env vars (AIBADGR_API_KEY, AIBADGR_BASE_URL)
• docs/my-website/docs/providers/aibadgr.md: Complete documentation with tier models (basic, normal, premium)

Usage:

import litellm
import os

os.environ["AIBADGR_API_KEY"] = "your-key"

response = litellm.completion(
    model="aibadgr/premium",
    messages=[{"role": "user", "content": "Hello"}]
)

Testing:
Existing test suite covers JSON providers. tests/test_litellm/llms/openai_like/test_json_providers.py includes AI Badgr test cases for provider resolution, tier models, and configuration loading.

Original prompt

Universal PR Prompt — Add AI Badgr (OpenAI-compatible) with minimal risk

You are adding AI Badgr as an optional OpenAI-compatible provider to THIS repo.

Goals

• Touch as few files as possible (1–3, hard max 5).
• Follow THIS repo’s existing provider/preset/docs pattern exactly.
• Keep the change optional, low-risk, and mergeable.

Hard constraints

• No new dependencies.
• No refactors. No renames. No formatting-only changes.
• Do not change defaults or core behavior.
• Do not add new runtime behavior (no background refresh, model catalog fetch, telemetry, caching, etc.).
• If the repo already has a provider/plugin system, integrate using it. Otherwise docs-only.

---

Step 0 — Choose exactly one integration mode

Inspect THIS repo and pick ONE path:

Path A — Provider registry / presets exist

Add AI Badgr as a selectable provider/preset using the same pattern as others.

Path B — Generic OpenAI-compatible endpoints exist

Add AI Badgr as a documented OpenAI-compatible endpoint (docs-only) using the repo’s existing mechanism.

Path C — UX labeling exists (optional)

If provider descriptions/tooltips exist, label it:

• “AI Badgr (Budget/Utility, OpenAI-compatible)”

---

Step 1 — Configuration details (use repo conventions)

• Display name: AI Badgr (or Path C label above)

• Base URL (default): https://aibadgr.com/api/v1

• Auth: Authorization: Bearer <API_KEY>

• Env vars (only if repo uses env vars):

  • AIBADGR_API_KEY
  • AIBADGR_BASE_URL (optional override)

If the repo uses a credential UI, follow existing credential patterns and name the field “AI Badgr API Key”.

Important: don’t hardcode secrets or add new config systems.

---

Step 2 — Model defaults (keep it simple)

In docs/examples/presets, default to tier names:

• basic
• normal
• premium (use this in examples)

If the repo already documents “advanced model selection”, optionally add:

• phi-3-mini (basic)
• mistral-7b (normal)
• llama3-8b-instruct (premium)

Add one line:

• “OpenAI-style model names are accepted and mapped automatically.”

---

Step 3 — Verification snippet (only if the repo already has them)

If other providers include a “verify” snippet, add one minimal snippet using model="premium".

---

Allowed files to change

• Provider registry/preset list (Path A)
• Provider docs (Path A/B)
• At most one example file (only if repo already has examples)

Hard max: 5 files.

---

Process requirements

• Follow THIS repo’s contribution guidelines.
• Run the repo’s minimal checks (tests/lint/docs checks) relevant to the changes.
• Ensure CI passes.

---

Output format at the end

• List files changed + reason per file

• Confirm:

  • followed repo patterns
  • no defaults changed
  • no new deps/runtime behavior
  • tier names used by default
  • premium used in examples
  • applied Budget/Utility label if Path C

---

What I changed (so you know why it’s better)

• Removed “maps automatically” being too absolute unless you know the mapping exists in your backend (maintainers hate unverifiable claims).
• Softened “AIBADGR_BASE_URL required” → “optional override” (more standard).
• Made Path C explicitly optional and dependent on repo having tooltips/descriptions.
• Reduced “test it all works” to “run minimal checks relevant to changes” (otherwise AI will claim it ran tests it didn’t).

---

One more practical tip

For each repo, add a one-line “why this belongs” that isn’t salesy, e.g.:

• “Adds another OpenAI-compatible endpoint option for users who want a budget provider.”

Maintainers merge “options,” not ads.

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.