feat: add OrcaRouter provider#311
Open
zhenjunchen-png wants to merge 1 commit into
Open
Conversation
Add OrcaRouter (https://www.orcarouter.ai), an OpenAI-compatible LLM router, following the existing cmd/aihubmix generator pattern. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Add OrcaRouter as a new OpenAI-compatible LLM router.
Disclosure: I'm an engineer on the OrcaRouter team.
Provider Details
https://api.orcarouter.ai/v1ORCAROUTER_API_KEYopenai-compat(OpenAI-compatible chat completions, streaming, function calling)Models
111 models, including the
orcarouter/autoadaptive router. A sample:orcarouter/autoanthropic/claude-opus-4.8openai/gpt-5.5google/gemini-3.5-flashdeepseek/deepseek-v4-proqwen/qwen3.7-maxChanges
cmd/orcarouter/main.go, a generator that builds the model list fromOrcaRouter's public model and pricing data, following the existing
cmd/aihubmixpattern.internal/providers/configs/orcarouter.json.internal/providers/providers.goand add theInferenceProviderOrcaRouterconstant inpkg/catwalk/provider.go.gen:orcaroutertoTaskfile.yamland an OrcaRouter step to theupdate.ymlworkflow.Reasoning is advertised for Gemini, Grok, Qwen, DeepSeek, MiniMax, Kimi, and
z.ai models. OpenAI and Anthropic models are served for chat and tool calls but
not reasoning: on chat completions OpenAI rejects
toolstogether withreasoning_effort(it requires the responses endpoint) and Anthropic uses itsown thinking field.
Validation
go build ./...,go vet ./...,go test ./...across vendors, the auto router, and error paths) and ran Crush against the
local catalog with OpenAI, Anthropic, and Gemini models all responding.
Usage
export ORCAROUTER_API_KEY=your-key-here crush --model orcarouter/anthropic/claude-opus-4.8