Switch to Ollama Cloud (OpenAI-compatible /v1 endpoint) for plumbing smoke test

config.yaml

@@ -1,26 +1,22 @@
 # LiteLLM config for inference.coop
-# MVP config: Tinfoil only (TEE-protected, privacy-first)
-# Two models: DeepSeek V4 Flash (default, best agents) + GPT-OSS 120B (cheapest)
-# Provider key set via environment variable or Admin UI
+# TEST config: Ollama Cloud (existing subscription) — plumbing smoke test
+# Uses Ollama Cloud's OpenAI-compatible endpoint (/v1) via LiteLLM's openai/ provider.
+# Provider key set via environment variable OLLAMA_API_KEY.
 
 model_list:
   # DeepSeek V4 Flash — default model
-  # Best agent performance (Terminal Bench 2.1: 82.7, DeepSWE: 54.4)
-  # MoE, 1M context, tool calling, $0.30/$0.70 per M tokens
   - model_name: deepseek-v4-flash
     litellm_params:
-      model: openai/deepseek-v4-flash
-      api_base: https://api.tinfoil.sh/v1
-      api_key: os.environ/TINFOIL_API_KEY
+      model: openai/deepseek-v4-flash:0731
+      api_base: https://ollama.com/v1
+      api_key: os.environ/OLLAMA_API_KEY
 
   # GPT-OSS 120B — lightweight fallback
-  # Built for agentic workflows, web search + code execution
-  # Apache 2.0, $0.15/$0.60 per M tokens
   - model_name: gpt-oss-120b
     litellm_params:
-      model: openai/gpt-oss-120b
-      api_base: https://api.tinfoil.sh/v1
-      api_key: os.environ/TINFOIL_API_KEY
+      model: openai/gpt-oss:120b
+      api_base: https://ollama.com/v1
+      api_key: os.environ/OLLAMA_API_KEY
 
 # Fallback: if DeepSeek is down, use GPT-OSS
 router_settings:
@@ -39,4 +35,4 @@ litellm_settings:
   salt_key: os.environ/LITELLM_SALT_KEY
   drop_params: true
   num_threads: 4
-  request_timeout: 30
+  request_timeout: 30