Skip to content

feat(providers): add Featherless as LLM provider#2336

Open
ArEnSc wants to merge 2 commits intoNousResearch:mainfrom
ArEnSc:feat/add-featherless-provider
Open

feat(providers): add Featherless as LLM provider#2336
ArEnSc wants to merge 2 commits intoNousResearch:mainfrom
ArEnSc:feat/add-featherless-provider

Conversation

@ArEnSc
Copy link
Copy Markdown

@ArEnSc ArEnSc commented Mar 21, 2026

Summary

  • Add Featherless as a new LLM provider with OpenAI-compatible API (https://api.featherless.ai/v1)
  • Featherless provides serverless inference for 30,000+ open-source HuggingFace models with flat-rate unlimited token pricing
  • Uses MiniMaxAI/MiniMax-M2.5 as default auxiliary model

Changes

  • hermes_cli/auth.py — Register featherless in PROVIDER_REGISTRY
  • hermes_cli/models.py — Add model catalog, provider label, and featherless-ai alias
  • agent/auxiliary_client.py — Add default auxiliary model for side tasks
  • .env.example — Add FEATHERLESS_API_KEY documentation
  • tests/test_api_key_providers.py — Add unit tests for registration, env vars, base URL, and provider resolution

Test plan

  • Verified provider registration imports correctly
  • Verified model catalog, label, and alias resolve correctly
  • Smoke tested live API call with NousResearch/Hermes-3-Llama-3.1-8B — response returned successfully
  • Run full test suite (pytest tests/test_api_key_providers.py)

🤖 Generated with Claude Code

ArEnSc and others added 2 commits March 21, 2026 12:39
Featherless provides serverless inference for 30,000+ open-source
HuggingFace models via an OpenAI-compatible API with flat-rate pricing.

- Register provider in PROVIDER_REGISTRY (auth.py)
- Add model catalog, label, and alias (models.py)
- Set MiniMax-M2.5 as default auxiliary model (auxiliary_client.py)
- Add env var documentation (.env.example)
- Add unit tests for registration, env vars, and resolution

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The alias map in resolve_provider (auth.py) is separate from
_PROVIDER_ALIASES in models.py — add featherless-ai there too.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant