Skip to content

Switchyard lets LLM applications route traffic across models and providers while preserving native OpenAI and Anthropic API compatibility - enabling flexible model selection, benchmarking, and cost/performance optimization.

Rust

chore: dev-server change decisions model (#1136)

Graham King50dfa7e28042
diff --git a/dev-server/config.toml b/dev-server/config.toml
index 2229c2102..d3eba94dc 100644
--- a/dev-server/config.toml
+++ b/dev-server/config.toml
@@ -8,10 +8,10 @@ base_url = "https://inference-api.nvidia.com/v1"
api_key_env = "NVIDIA_API_KEY"
[decision_clients.decisions]
-format = "system_one"
-endpoint = "https://openrouter.ai/api/alpha/decisions"
-api_key_env = "OPENROUTER_API_KEY"
-timeout_ms = 15000
+format = "openai_decisions"
+endpoint = "https://inference-api.nvidia.com/v1/decisions"
+api_key_env = "NVIDIA_API_KEY"
+timeout_ms = 5000
# The three models our algos use
@@ -26,7 +26,7 @@ llm_client = "inference_hub"
extra_body = { reasoning = { effort = "medium" } }
[decision_targets.classifier]
-id = "inception/mercury-decide:free"
+id = "openai/openai/gpt-6-luna"
decision_client = "decisions"
# Best route
diff --git a/dev-server/switchyard.service b/dev-server/switchyard.service
index dafefcf5d..51962a9a1 100644
--- a/dev-server/switchyard.service
+++ b/dev-server/switchyard.service
@@ -19,7 +19,6 @@ ExecStart=/usr/local/bin/switchyard-server \
Restart=on-failure
Environment="NVIDIA_API_KEY=<PUT-THE-KEY-HERE>"
-Environment="OPENROUTER_API_KEY=<PUT-THE-OTHER-KEY-HERE>"
PrivateDevices=yes
ProtectHome=yes