Skip to content

Feat/more openrouter models - #2

Merged
vrana055 merged 3 commits into
mainfrom
feat/more-openrouter-models
Jun 22, 2026
Merged

vrana055 merged 3 commits into
mainfrom
feat/more-openrouter-models

Conversation

@vrana055

Copy link
Copy Markdown
Contributor

No description provided.

vrana055 and others added 3 commits June 21, 2026 18:31
Add 5 new free models (gpt-oss-120b, qwen3-coder, nex-n2-pro,
gpt-oss-20b, nemotron-omni-30b-reasoning) verified against the
OpenRouter API. Inserted at high-priority slots so larger/stronger
models are tried before smaller fallbacks.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Refactor llm.py to route each model to its own OpenAI-compatible client
via "provider>model_id" format in the chain. No prefix = openrouter
(backward compat). Add Groq as second provider with 3 free models
(llama-3.3-70b-versatile, qwen3.6-27b, llama-3.1-8b-instant).

Adding a future provider requires: one dict entry in _provider_registry(),
two env vars in config.py, and prefixed model IDs in GENERATION_MODELS.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Extend multi-provider registry with Cerebras (wafer-scale ~2000 tok/s)
and Mistral AI (free-tier). Add 5 new models to fallback chain:
cerebras>llama-3.3-70b, cerebras>qwen-3-32b, cerebras>llama-3.1-8b,
mistral>mistral-small-latest, mistral>open-mistral-nemo.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
@vrana055
vrana055 merged commit 0b1d3ab into main Jun 22, 2026
1 of 4 checks passed
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