Add Ollama Cloud provider

Ollama's hosted service (https://ollama.com) exposes OpenAI-compatible
endpoints at /v1 with Bearer-token auth. Add :ollama_cloud as a dedicated
provider inheriting from the existing Ollama provider so chat, streaming,
media, and dynamic model listing all work unchanged, while correctly
reporting as remote, requiring an API key, and defaulting api_base to
https://ollama.com/v1.

Two class-level overrides are load-bearing:

  - `slug` returns "ollama_cloud" — the default "ollamacloud" would
    mismatch the :ollama_cloud registration symbol and break
    Model::Info#provider lookups.

  - `assume_models_exist?` returns true — cloud models are dynamic and
    not in the static registry; the existing Ollama provider gets the
    same shortcut via its `local?` flag, which OllamaCloud correctly
    returns false for.

Models.dev already catalogs Ollama Cloud under the key "ollama-cloud",
so the MODELS_DEV_PROVIDER_MAP entry wires its 37 models into the
shared registry with full metadata (context_window, max_output_tokens,
capabilities). Adding ollama_cloud_api_key to models.rake's
configure_from_env lets the maintainer's next `rake models:update`
populate the shipped models.json.

Verified live against the hosted API: /v1/models returns the expected
OpenAI list shape, sync and streaming chat both work on gpt-oss:120b,
ConfigurationError is raised when the key is missing, and
`RubyLLM.models.refresh!` populates 38 entries (37 from models.dev +
1 additional from the live provider listing).

8 VCR cassettes recorded (4 basic chat, 1 streaming, 3 thinking). Two
reasoning-model quirks match the existing ollama/qwen3 skip pattern:
system-prompt replacement and streaming-vs-sync token count drift.

Resolves #740.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Author: dalton-cole

.env.example

@@ -13,6 +13,7 @@ GPUSTACK_API_BASE=http://localhost:11444/v1
 GPUSTACK_API_KEY=$(op read "op://RubyLLM/GPUStack/credential")
 MISTRAL_API_KEY=$(op read "op://RubyLLM/Mistral/credential")
 OLLAMA_API_BASE=http://localhost:11434/v1
+OLLAMA_CLOUD_API_KEY=$(op read "op://RubyLLM/Ollama Cloud/credential")
 OPENAI_API_KEY=$(op read "op://RubyLLM/OpenAI/credential")
 OPENROUTER_API_KEY=$(op read "op://RubyLLM/OpenRouter/credential")
 PERPLEXITY_API_KEY=$(op read "op://RubyLLM/Perplexity/credential")

README.md

@@ -140,7 +140,7 @@ response = chat.with_schema(ProductSchema).ask "Analyze this product", with: "pr
 * **Async:** Fiber-based concurrency
 * **Model registry:** 800+ models with capability detection and pricing
 * **Extended thinking:** Control, view, and persist model deliberation
-* **Providers:** OpenAI, xAI, Anthropic, Gemini, VertexAI, Bedrock, DeepSeek, Mistral, Ollama, OpenRouter, Perplexity, GPUStack, and any OpenAI-compatible API
+* **Providers:** OpenAI, xAI, Anthropic, Gemini, VertexAI, Bedrock, DeepSeek, Mistral, Ollama, Ollama Cloud, OpenRouter, Perplexity, GPUStack, and any OpenAI-compatible API
 
 ## Installation
 

docs/_getting_started/configuration.md

@@ -83,6 +83,10 @@ RubyLLM.configure do |config|
   config.ollama_api_base = 'http://localhost:11434/v1'
   config.ollama_api_key = ENV['OLLAMA_API_KEY'] # Available in v1.13.0+ (optional for authenticated/remote Ollama endpoints)
 
+  # Ollama Cloud
+  config.ollama_cloud_api_key = ENV['OLLAMA_CLOUD_API_KEY']   # Required. Keys: https://ollama.com/settings/keys
+  config.ollama_cloud_api_base = ENV['OLLAMA_CLOUD_API_BASE'] # Optional. Defaults to https://ollama.com/v1
+
   # OpenAI
   config.openai_api_key = ENV['OPENAI_API_KEY']
   config.openai_api_base = ENV['OPENAI_API_BASE'] # Optional custom OpenAI-compatible endpoint
@@ -166,6 +170,28 @@ end
 
 By default, RubyLLM uses the 'developer' role (matching OpenAI's current API). Set `openai_use_system_role` to true for compatibility with servers that still expect 'system'.
 
+### Ollama Cloud
+
+Ollama's hosted service exposes OpenAI-compatible endpoints at `https://ollama.com/v1` with Bearer-token auth. Keys are issued at [ollama.com/settings/keys](https://ollama.com/settings/keys).
+
+```ruby
+RubyLLM.configure do |config|
+  config.ollama_cloud_api_key = ENV['OLLAMA_CLOUD_API_KEY']
+end
+
+chat = RubyLLM.chat(
+  model: 'gpt-oss:120b',
+  provider: :ollama_cloud,
+  assume_model_exists: true
+)
+chat.ask('Hello from the cloud')
+```
+
+Cloud-capable models include `gpt-oss:120b`, `gpt-oss:120b-cloud`, `qwen3-coder:480b-cloud`, and `deepseek-v3.1:671b-cloud`. Models are discovered dynamically via `/v1/models`; pass `assume_model_exists: true` until you run `RubyLLM.models.refresh!`.
+
+> Ollama Cloud is billed by subscription tier (Free / Pro $20/mo / Max $100/mo), not per-token — so `Message#input_tokens` and `Message#output_tokens` are reported but `Model::Info#pricing` will be empty. See [ollama.com/pricing](https://ollama.com/pricing) for current tiers.
+{: .note }
+
 ### Gemini API Versions
 {: .d-inline-block }
 
@@ -484,6 +510,10 @@ RubyLLM.configure do |config|
   config.ollama_api_base = String
   config.ollama_api_key = String  # v1.13.0+
 
+  # Ollama Cloud
+  config.ollama_cloud_api_key = String
+  config.ollama_cloud_api_base = String
+
   # OpenAI
   config.openai_api_key = String
   config.openai_api_base = String

docs/index.md

@@ -214,7 +214,7 @@ response = chat.with_schema(ProductSchema).ask "Analyze this product", with: "pr
 * **Async:** Fiber-based concurrency
 * **Model registry:** 800+ models with capability detection and pricing
 * **Extended thinking:** Control, view, and persist model deliberation
-* **Providers:** OpenAI, xAI, Anthropic, Gemini, VertexAI, Bedrock, DeepSeek, Mistral, Ollama, OpenRouter, Perplexity, GPUStack, and any OpenAI-compatible API
+* **Providers:** OpenAI, xAI, Anthropic, Gemini, VertexAI, Bedrock, DeepSeek, Mistral, Ollama, Ollama Cloud, OpenRouter, Perplexity, GPUStack, and any OpenAI-compatible API
 
 ## Installation
 

lib/ruby_llm.rb

@@ -101,6 +101,7 @@ def logger
 RubyLLM::Provider.register :gpustack, RubyLLM::Providers::GPUStack
 RubyLLM::Provider.register :mistral, RubyLLM::Providers::Mistral
 RubyLLM::Provider.register :ollama, RubyLLM::Providers::Ollama
+RubyLLM::Provider.register :ollama_cloud, RubyLLM::Providers::OllamaCloud
 RubyLLM::Provider.register :openai, RubyLLM::Providers::OpenAI
 RubyLLM::Provider.register :openrouter, RubyLLM::Providers::OpenRouter
 RubyLLM::Provider.register :perplexity, RubyLLM::Providers::Perplexity

lib/ruby_llm/models.rb

@@ -13,6 +13,7 @@ class Models
       'amazon-bedrock' => 'bedrock',
       'deepseek' => 'deepseek',
       'mistral' => 'mistral',
+      'ollama-cloud' => 'ollama_cloud',
       'openrouter' => 'openrouter',
       'perplexity' => 'perplexity'
     }.freeze

lib/ruby_llm/providers/ollama_cloud.rb

@@ -0,0 +1,42 @@
+# frozen_string_literal: true
+
+module RubyLLM
+  module Providers
+    # Ollama Cloud API integration.
+    class OllamaCloud < Ollama
+      def api_base
+        @config.ollama_cloud_api_base || 'https://ollama.com/v1'
+      end
+
+      def headers
+        { 'Authorization' => "Bearer #{@config.ollama_cloud_api_key}" }
+      end
+
+      class << self
+        def slug
+          'ollama_cloud'
+        end
+
+        def configuration_options
+          %i[ollama_cloud_api_base ollama_cloud_api_key]
+        end
+
+        def configuration_requirements
+          %i[ollama_cloud_api_key]
+        end
+
+        def local?
+          false
+        end
+
+        def assume_models_exist?
+          true
+        end
+
+        def capabilities
+          Ollama::Capabilities
+        end
+      end
+    end
+  end
+end

lib/tasks/models.rake

@@ -45,6 +45,7 @@ def configure_from_env
     config.deepseek_api_key = ENV.fetch('DEEPSEEK_API_KEY', nil)
     config.gemini_api_key = ENV.fetch('GEMINI_API_KEY', nil)
     config.mistral_api_key = ENV.fetch('MISTRAL_API_KEY', nil)
+    config.ollama_cloud_api_key = ENV.fetch('OLLAMA_CLOUD_API_KEY', nil)
     config.openai_api_key = ENV.fetch('OPENAI_API_KEY', nil)
     config.openrouter_api_key = ENV.fetch('OPENROUTER_API_KEY', nil)
     config.perplexity_api_key = ENV.fetch('PERPLEXITY_API_KEY', nil)