feat: add CloudGPU as an OpenAI-compatible provider alias

agent-schema.json

@@ -224,7 +224,7 @@
       "properties": {
         "provider": {
           "type": "string",
-          "description": "The underlying provider type. Defaults to \"openai\" when not set. Supported values: openai, anthropic, google, amazon-bedrock, dmr, and any built-in alias (requesty, openrouter, opper, runcrate, api-route, orcarouter, atlascloud, azure, xai, ollama, mistral, baseten, daoxe, ovhcloud, groq, fireworks-ai, deepseek, cerebras, togetherai, huggingface, moonshotai, vercel, cloudflare-workers-ai, cloudflare-ai-gateway, nvidia, github-copilot, chatgpt, opencode, opencode-go, etc.).",
+          "description": "The underlying provider type. Defaults to \"openai\" when not set. Supported values: openai, anthropic, google, amazon-bedrock, dmr, and any built-in alias (requesty, openrouter, opper, runcrate, cloudgpu, api-route, orcarouter, atlascloud, azure, xai, ollama, mistral, baseten, daoxe, ovhcloud, groq, fireworks-ai, deepseek, cerebras, togetherai, huggingface, moonshotai, vercel, cloudflare-workers-ai, cloudflare-ai-gateway, nvidia, github-copilot, chatgpt, opencode, opencode-go, etc.).",
           "examples": [
             "openai",
             "anthropic",

docs/configuration/models/index.md

@@ -18,7 +18,7 @@ models:
     first_available: [list] # Optional: candidate model refs, tried in order by available credentials.
                             # Mutually exclusive with other model settings.
     provider: string # Required unless using first_available. One of: openai, anthropic, google, amazon-bedrock,
-                     # dmr, mistral, xai, nebius, nvidia, minimax, baseten, daoxe, ovhcloud, groq, fireworks-ai, deepseek, cerebras, togetherai, huggingface, moonshotai, vercel, cloudflare-workers-ai, cloudflare-ai-gateway, requesty, openrouter, opper, runcrate, api-route, orcarouter, atlascloud,
+                     # dmr, mistral, xai, nebius, nvidia, minimax, baseten, daoxe, ovhcloud, groq, fireworks-ai, deepseek, cerebras, togetherai, huggingface, moonshotai, vercel, cloudflare-workers-ai, cloudflare-ai-gateway, requesty, openrouter, opper, runcrate, cloudgpu, api-route, orcarouter, atlascloud,
                      # azure, ollama, github-copilot, or a named provider defined
                      # under the top-level `providers:` section.
     model: string # Required: model identifier
@@ -60,7 +60,7 @@ models:
 | Property              | Type       | Required | Description                                                                           |
 | --------------------- | ---------- | -------- | ------------------------------------------------------------------------------------- |
 | `first_available`     | array      | ✗        | Candidate model references tried in order; selects the first whose credentials are configured. Mutually exclusive with other model settings. |
-| `provider`            | string     | ✓/✗      | Required for regular model definitions; omitted for `first_available` selectors. Provider: `openai`, `anthropic`, `google`, `amazon-bedrock`, `dmr`, `mistral`, `xai`, `nebius`, `nvidia`, `minimax`, `baseten`, `daoxe`, `ovhcloud`, `groq`, `fireworks-ai`, `deepseek`, `cerebras`, `togetherai`, `huggingface`, `moonshotai`, `vercel`, `cloudflare-workers-ai`, `cloudflare-ai-gateway`, `requesty`, `openrouter`, `opper`, `runcrate`, `api-route`, `orcarouter`, `atlascloud`, `azure`, `ollama`, `github-copilot`, `chatgpt`, `opencode`, `opencode-go`, or any [named provider](../../providers/custom/index.md). |
+| `provider`            | string     | ✓/✗      | Required for regular model definitions; omitted for `first_available` selectors. Provider: `openai`, `anthropic`, `google`, `amazon-bedrock`, `dmr`, `mistral`, `xai`, `nebius`, `nvidia`, `minimax`, `baseten`, `daoxe`, `ovhcloud`, `groq`, `fireworks-ai`, `deepseek`, `cerebras`, `togetherai`, `huggingface`, `moonshotai`, `vercel`, `cloudflare-workers-ai`, `cloudflare-ai-gateway`, `requesty`, `openrouter`, `opper`, `runcrate`, `cloudgpu`, `api-route`, `orcarouter`, `atlascloud`, `azure`, `ollama`, `github-copilot`, `chatgpt`, `opencode`, `opencode-go`, or any [named provider](../../providers/custom/index.md). |
 | `model`               | string     | ✓/✗      | Required for regular model definitions; omitted for `first_available` selectors. Model name (e.g., `gpt-4o`, `claude-sonnet-4-5`, `gemini-3.5-flash`) |
 | `description`         | string     | ✗        | Informational, human-readable summary of the model's purpose or strengths (e.g., "fast and cheap, good for summaries"). Not sent to the model. Can be combined with `first_available` (a selector's description is kept when it resolves). |
 | `temperature`         | float      | ✗        | Sampling randomness. Range is provider-dependent — typically `0.0–2.0` (Anthropic caps at `1.0`). `0.0` is deterministic. |
@@ -558,7 +558,7 @@ for a complete local-server configuration.
 ## Custom HTTP Headers
 
 For OpenAI-compatible providers (`openai`, `github-copilot`, `mistral`, `xai`,
-`nebius`, `nvidia`, `minimax`, `baseten`, `daoxe`, `ovhcloud`, `groq`, `fireworks-ai`, `deepseek`, `cerebras`, `togetherai`, `huggingface`, `moonshotai`, `vercel`, `cloudflare-workers-ai`, `cloudflare-ai-gateway`, `requesty`, `openrouter`, `opper`, `runcrate`, `orcarouter`, `atlascloud`, `ollama`, and any custom provider using the OpenAI API),
+`nebius`, `nvidia`, `minimax`, `baseten`, `daoxe`, `ovhcloud`, `groq`, `fireworks-ai`, `deepseek`, `cerebras`, `togetherai`, `huggingface`, `moonshotai`, `vercel`, `cloudflare-workers-ai`, `cloudflare-ai-gateway`, `requesty`, `openrouter`, `opper`, `runcrate`, `cloudgpu`, `orcarouter`, `atlascloud`, `ollama`, and any custom provider using the OpenAI API),
 `provider_opts.http_headers` adds arbitrary HTTP headers to every outgoing
 request:
 

docs/data/nav.yml

@@ -237,6 +237,8 @@
       url: /providers/ovhcloud/
     - title: Runcrate
       url: /providers/runcrate/
+    - title: CloudGPU
+      url: /providers/cloudgpu/
     - title: Together AI
       url: /providers/together/
     - title: Vercel AI Gateway

docs/providers/cloudgpu/index.md

@@ -0,0 +1,110 @@
+---
+title: "CloudGPU"
+description: "Use CloudGPU models with Docker Agent."
+keywords: docker agent, ai agents, model providers, llm, cloudgpu
+weight: 25
+canonical: https://docs.docker.com/ai/docker-agent/providers/cloudgpu/
+---
+
+_Use CloudGPU models with Docker Agent._
+
+## Overview
+
+[CloudGPU](https://cloudgpu.app/api?utm_source=github&utm_medium=directory&utm_campaign=202611_seg04_codingagent&utm_content=github_docker_agent_provider)
+rents cloud GPUs and serves DeepSeek, Kimi, GLM, MiniMax and other models
+through an OpenAI-compatible API. Docker Agent includes built-in support for
+CloudGPU as an alias provider.
+
+## Setup
+
+1. Create an API key in the [CloudGPU console](https://cloudgpu.app/)
+2. Set the environment variable:
+
+   ```bash
+   export CLOUDGPU_API_KEY=your-api-key
+   ```
+
+## Usage
+
+### Inline Syntax
+
+Use a CloudGPU model by prefixing its model ID with `cloudgpu/`:
+
+```yaml
+agents:
+  root:
+    model: cloudgpu/deepseek-v4.1-flash
+    description: Assistant using CloudGPU
+    instruction: You are a helpful assistant.
+```
+
+### Named Model
+
+For more control over parameters:
+
+```yaml
+models:
+  cloudgpu_kimi:
+    provider: cloudgpu
+    model: kimi-k2.6
+    temperature: 0.7
+    max_tokens: 8192
+
+agents:
+  root:
+    model: cloudgpu_kimi
+    description: Assistant using CloudGPU
+    instruction: You are a helpful assistant.
+```
+
+## Available Models
+
+Some chat model IDs CloudGPU serves:
+
+| Model ID | Model |
+| --- | --- |
+| `deepseek-v4.1-flash` | DeepSeek V4.1 Flash (default) |
+| `kimi-k2.6` | Kimi K2.6 |
+| `glm-5.3-flash` | GLM-5.3 Flash |
+| `minimax-m3` | MiniMax M3 |
+
+Model IDs are case-sensitive. The live [`/v1/models` endpoint](https://cloudgpu.app/v1/models)
+lists current IDs, and [the API docs](https://cloudgpu.app/api/docs) describe
+the endpoints.
+
+CloudGPU is not in the models.dev catalog. Docker Agent does not import metadata
+from CloudGPU's model list, so it has no catalog-derived pricing or context limits
+for these models. Attachment capabilities conservatively fall back to text-only;
+image and PDF attachments are dropped, with a warning logged once per model.
+
+If you have verified that your chosen model accepts a particular attachment
+type, set `models.<name>.capabilities` to override detection. See
+[Attachment Capability Overrides](../../configuration/models/index.md#attachment-capability-overrides)
+for all supported flags.
+
+## How It Works
+
+CloudGPU is implemented as a built-in alias in Docker Agent:
+
+- **API Type:** OpenAI-compatible Chat Completions (`openai_chatcompletions`)
+- **Base URL:** `https://cloudgpu.app/v1`
+- **Token Variable:** `CLOUDGPU_API_KEY`
+
+Some model chat templates reject multiple leading system messages, so Docker
+Agent coalesces its per-source system messages into one for CloudGPU.
+
+## Example: Code Assistant
+
+```yaml
+agents:
+  coder:
+    model: cloudgpu/deepseek-v4.1-flash
+    description: Code assistant using CloudGPU
+    instruction: |
+      You are an expert programmer.
+      Write clean code and follow language best practices.
+    toolsets:
+      - type: filesystem
+      - type: shell
+      - type: think
+```

docs/providers/overview/index.md

@@ -72,6 +72,7 @@ Use this table to find a built-in provider's config key and authentication metho
 | OrcaRouter     | `orcarouter`     | `ORCAROUTER_API_KEY`                |
 | Atlas Cloud    | [`atlascloud`](../atlascloud/index.md) | `ATLASCLOUD_API_KEY`                |
 | Runcrate       | [`runcrate`](../runcrate/index.md) | `RUNCRATE_API_KEY`                  |
+| CloudGPU       | [`cloudgpu`](../cloudgpu/index.md) | `CLOUDGPU_API_KEY`                  |
 | Azure OpenAI   | `azure`          | `AZURE_API_KEY` + `base_url` (override the key with `token_key`) |
 | [Ollama](../local/index.md) | `ollama` | None (local; optional `base_url`) |
 | GitHub Copilot | `github-copilot` | `GITHUB_TOKEN` or `GH_TOKEN` (PAT with `copilot` scope) |

examples/README.md

@@ -214,6 +214,7 @@ remote MCP endpoints.
 | [`orcarouter.yaml`](orcarouter.yaml) | OrcaRouter multi-model gateway provider. |
 | [`ovhcloud.yaml`](ovhcloud.yaml) | OVHcloud AI Endpoints provider. |
 | [`runcrate.yaml`](runcrate.yaml) | Runcrate open-model host (`runcrate`). |
+| [`cloudgpu.yaml`](cloudgpu.yaml) | CloudGPU OpenAI-compatible model API (`cloudgpu`). |
 | [`groq.yaml`](groq.yaml) | Groq fast-inference provider. |
 | [`fireworks.yaml`](fireworks.yaml) | Fireworks AI open-model inference (`fireworks-ai`). |
 | [`deepseek.yaml`](deepseek.yaml) | DeepSeek chat and reasoning provider. |

examples/cloudgpu.yaml

@@ -0,0 +1,17 @@
+# yaml-language-server: $schema=../agent-schema.json
+
+models:
+  cloudgpu_model:
+    provider: cloudgpu
+    model: deepseek-v4.1-flash
+
+agents:
+  root:
+    model: cloudgpu_model
+    description: Assistant using CloudGPU
+    instruction: |
+      You are a helpful assistant.
+    toolsets:
+      - type: filesystem
+      - type: shell
+      - type: think

pkg/config/auto.go

@@ -67,6 +67,7 @@ var cloudProviders = []providerConfig{
 	{"atlascloud", []string{"ATLASCLOUD_API_KEY"}, "ATLASCLOUD_API_KEY", "ATLASCLOUD_API_KEY"},
 	{"opper", []string{"OPPER_API_KEY"}, "OPPER_API_KEY", "OPPER_API_KEY"},
 	{"runcrate", []string{"RUNCRATE_API_KEY"}, "RUNCRATE_API_KEY", "RUNCRATE_API_KEY"},
+	{"cloudgpu", []string{"CLOUDGPU_API_KEY"}, "CLOUDGPU_API_KEY", "CLOUDGPU_API_KEY"},
 	{"baseten", []string{"BASETEN_API_KEY"}, "BASETEN_API_KEY", "BASETEN_API_KEY"},
 	{"ovhcloud", []string{"OVH_AI_ENDPOINTS_ACCESS_TOKEN"}, "OVH_AI_ENDPOINTS_ACCESS_TOKEN", "OVH_AI_ENDPOINTS_ACCESS_TOKEN"},
 	{"daoxe", []string{"DAOXE_API_KEY"}, "DAOXE_API_KEY", "DAOXE_API_KEY"},
@@ -169,6 +170,7 @@ var DefaultModels = map[string]string{
 	"atlascloud":     "qwen/qwen3.8-max",
 	"opper":          "claude-sonnet-4-6",
 	"runcrate":       "deepseek-ai/DeepSeek-V4.1-Flash",
+	"cloudgpu":       "deepseek-v4.1-flash",
 	"baseten":        "deepseek-ai/DeepSeek-V4-Pro-0813",
 	"ovhcloud":       "Qwen3.5-397B-A17B",
 	"daoxe":          "claude-sonnet-4-6",

pkg/config/auto_test.go

@@ -109,6 +109,13 @@ func TestAvailableProviders_NoGateway(t *testing.T) {
 			},
 			expectedProvider: "runcrate",
 		},
+		{
+			name: "cloudgpu api key present",
+			envVars: map[string]string{
+				"CLOUDGPU_API_KEY": "test-key",
+			},
+			expectedProvider: "cloudgpu",
+		},
 		{
 			name: "baseten api key present",
 			envVars: map[string]string{
@@ -375,6 +382,15 @@ func TestAutoModelConfig(t *testing.T) {
 			expectedModel:     "deepseek-ai/DeepSeek-V4.1-Flash",
 			expectedMaxTokens: 32000,
 		},
+		{
+			name: "cloudgpu provider",
+			envVars: map[string]string{
+				"CLOUDGPU_API_KEY": "test-key",
+			},
+			expectedProvider:  "cloudgpu",
+			expectedModel:     "deepseek-v4.1-flash",
+			expectedMaxTokens: 32000,
+		},
 		{
 			name: "orcarouter provider",
 			envVars: map[string]string{

pkg/config/examples_test.go

@@ -32,6 +32,7 @@ var modelsDevAbsentProviders = map[string]bool{
 	"api-route":             true, // API Route model IDs come from its own /v1/models endpoint, not models.dev.
 	"atlascloud":            true, // Atlas Cloud model IDs come from its own /v1/models endpoint, not models.dev.
 	"runcrate":              true, // Runcrate model IDs come from its own /v1/models endpoint, not models.dev.
+	"cloudgpu":              true, // CloudGPU model IDs come from its own /v1/models endpoint, not models.dev.
 	"dmr":                   true, // Docker Model Runner (local, not in catalog)
 	"cloudflare-workers-ai": true, // example uses an @cf/... model id not present in the models.dev snapshot (only variant ids like -fp8 are listed)
 	"cloudflare-ai-gateway": true, // multi-provider router; example model ids use the gateway's provider/model form, not guaranteed to match a models.dev id

pkg/model/provider/aliases.go

@@ -75,6 +75,11 @@ var Aliases = map[string]Alias{
 		BaseURL:     "https://api.runcrate.ai/v1",
 		TokenEnvVar: "RUNCRATE_API_KEY",
 	},
+	"cloudgpu": {
+		APIType:     "openai",
+		BaseURL:     "https://cloudgpu.app/v1",
+		TokenEnvVar: "CLOUDGPU_API_KEY",
+	},
 	"orcarouter": {
 		APIType:     "openai",
 		BaseURL:     "https://api.orcarouter.ai/v1",

pkg/model/provider/aliases_test.go

@@ -43,6 +43,7 @@ func TestCatalogAliases(t *testing.T) {
 		"orcarouter":   {APIType: "openai", BaseURL: "https://api.orcarouter.ai/v1", TokenEnvVar: "ORCAROUTER_API_KEY"},
 		"atlascloud":   {APIType: "openai", BaseURL: "https://api.atlascloud.ai/v1", TokenEnvVar: "ATLASCLOUD_API_KEY"},
 		"runcrate":     {APIType: "openai", BaseURL: "https://api.runcrate.ai/v1", TokenEnvVar: "RUNCRATE_API_KEY"},
+		"cloudgpu":     {APIType: "openai", BaseURL: "https://cloudgpu.app/v1", TokenEnvVar: "CLOUDGPU_API_KEY"},
 		"opper":        {APIType: "openai", BaseURL: "https://api.opper.ai/v3/compat", TokenEnvVar: "OPPER_API_KEY"},
 		"baseten":      {APIType: "openai", BaseURL: "https://inference.baseten.co/v1", TokenEnvVar: "BASETEN_API_KEY"},
 		"daoxe":        {APIType: "openai", BaseURL: "https://api.daoxe.com/v1", TokenEnvVar: "DAOXE_API_KEY"},
@@ -135,6 +136,7 @@ func TestProviderIDsMatchEmbeddedCatalog(t *testing.T) {
 		"atlascloud": "model IDs come from Atlas Cloud's /v1/models endpoint, not models.dev",
 		"chatgpt":    "subscription backend, distinct from OpenAI",
 		"runcrate":   "model IDs come from Runcrate's /v1/models endpoint, not models.dev",
+		"cloudgpu":   "model IDs come from CloudGPU's /v1/models endpoint, not models.dev",
 		"dmr":        "local Docker Model Runner",
 		"ollama":     "local Ollama server",
 	}

pkg/model/provider/openai/client.go

@@ -283,6 +283,7 @@ var openModelHostProviders = map[string]bool{
 	"openrouter":   true,
 	"opper":        true,
 	"runcrate":     true,
+	"cloudgpu":     true,
 	"orcarouter":   true,
 	"nebius":       true,
 	"nvidia":       true,

pkg/model/provider/openai/system_message_merge_test.go

@@ -75,6 +75,16 @@ func TestRuncrate_MergesConsecutiveSystemMessages(t *testing.T) {
 	})
 }
 
+func TestCloudGPU_MergesConsecutiveSystemMessages(t *testing.T) {
+	t.Parallel()
+
+	assertMergesConsecutiveMessages(t, &latest.ModelConfig{
+		Provider: "cloudgpu",
+		Model:    "deepseek-v4.1-flash",
+		TokenKey: "MY_TOKEN",
+	})
+}
+
 func TestBaseten_MergesConsecutiveSystemMessages(t *testing.T) {
 	t.Parallel()
 
@@ -204,6 +214,7 @@ func TestShouldMergeConsecutiveMessages_Gating(t *testing.T) {
 		{"daoxe, no base_url", &latest.ModelConfig{Provider: "daoxe", Model: "claude-sonnet-4-6"}, true},
 		{"opper, no base_url", &latest.ModelConfig{Provider: "opper", Model: "claude-sonnet-4-6"}, true},
 		{"runcrate, no base_url", &latest.ModelConfig{Provider: "runcrate", Model: "deepseek-ai/DeepSeek-V4.1-Flash"}, true},
+		{"cloudgpu, no base_url", &latest.ModelConfig{Provider: "cloudgpu", Model: "deepseek-v4.1-flash"}, true},
 		{"ovhcloud", &latest.ModelConfig{Provider: "ovhcloud", Model: "Qwen3.5-397B-A17B"}, true},
 		{"open-model host alias cerebras", &latest.ModelConfig{Provider: "cerebras", Model: "qwen-3-coder-480b"}, true},
 		{"open-model host fireworks", &latest.ModelConfig{Provider: "fireworks", Model: "accounts/fireworks/models/kimi-k3"}, true},

pkg/model/provider/openai_alias_providers_test.go

@@ -89,6 +89,14 @@ var openAIAliasProviders = []openAIAliasProvider{
 		greeting:             "Hello from Runcrate",
 		mergesSystemMessages: true,
 	},
+	{
+		provider:             "cloudgpu",
+		envVar:               "CLOUDGPU_API_KEY",
+		testKey:              "test-cloudgpu-key",
+		model:                "deepseek-v4.1-flash",
+		greeting:             "Hello from CloudGPU",
+		mergesSystemMessages: true,
+	},
 	{
 		provider: "deepseek",
 		envVar:   "DEEPSEEK_API_KEY",