Files
ollama-ollama/cmd/config/config.go
Devon Rifkin 8207e55ec7 don't require pulling stubs for cloud models (#14574)
* don't require pulling stubs for cloud models

This is a first in a series of PRs that will better integrate Ollama's
cloud into the API and CLI. Previously we used to have a layer of
indirection where you'd first have to pull a "stub" model that contains
a reference to a cloud model. With this change, you don't have to pull
first, you can just use a cloud model in various routes like `/api/chat`
and `/api/show`. This change respects
<https://github.com/ollama/ollama/pull/14221>, so if cloud is disabled,
these models won't be accessible.

There's also a new, simpler pass-through proxy that doesn't convert the
requests ahead of hitting the cloud models, which they themselves
already support various formats (e.g., `v1/chat/completions` or Open
Responses, etc.). This will help prevent issues caused by double
converting (e.g., `v1/chat/completions` converted to `api/chat` on the
client, then calling cloud and converting back to a
`v1/chat/completions` response instead of the cloud model handling the
original `v1/chat/completions` request first).

There's now a notion of "source tags", which can be mixed with existing
tags. So instead of having different formats like`gpt-oss:20b-cloud` vs.
`kimi-k2.5:cloud` (`-cloud` suffix vs. `:cloud`), you can now specify
cloud by simply appending `:cloud`. This PR doesn't change model
resolution yet, but sets us up to allow for things like omitting the
non-source tag, which would make something like `ollama run
gpt-oss:cloud` work the same way that `ollama run gpt-oss` already works
today.

More detailed changes:

- Added a shared model selector parser in `types/modelselector`:
  - supports `:cloud` and `:local`
  - accepts source tags in any position
  - supports legacy `:<tag>-cloud`
  - rejects conflicting source tags
- Integrated selector handling across server inference/show routes:
  - `GenerateHandler`, `ChatHandler`, `EmbedHandler`,
    `EmbeddingsHandler`, `ShowHandler`
- Added explicit-cloud passthrough proxy for ollama.com:
  - same-endpoint forwarding for `/api/*`, `/v1/*`, and `/v1/messages`
  - normalizes `model` (and `name` for `/api/show`) before forwarding
  - forwards request headers except hop-by-hop/proxy-managed headers
  - uses bounded response-header timeout
  - handles auth failures in a friendly way
- Preserved cloud-disable behavior (`OLLAMA_NO_CLOUD`)
- Updated create flow to support `FROM ...:cloud` model sources (though
  this flow uses the legacy proxy still, supporting Modelfile overrides
  is more complicated with the direct proxy approach)
- Updated CLI/TUI/config cloud detection to use shared selector logic
- Updated CLI preflight behavior so explicit cloud requests do not
  auto-pull local stubs

What's next?

- Cloud discovery/listing and cache-backed `ollama ls` / `/api/tags`
- Modelfile overlay support for virtual cloud models on OpenAI/Anthropic
  request families
- Recommender/default-selection behavior for ambiguous model families
- Fully remove the legacy flow

Fixes: https://github.com/ollama/ollama/issues/13801

* consolidate pull logic into confirmAndPull helper

pullIfNeeded and ShowOrPull shared identical confirm-and-pull logic.
Extract confirmAndPull to eliminate the duplication.

* skip local existence checks for cloud models

ModelExists and the TUI's modelExists both check the local model list,
which causes cloud models to appear missing. Return true early for
explicit cloud models so the TUI displays them beside the integration
name and skips re-prompting the model picker on relaunch.

* support optionally pulling stubs for newly-style names

We now normalize names like `<family>:<size>:cloud` into legacy-style
names like `<family>:<size>-cloud` for pulling and deleting (this also
supports stripping `:local`). Support for pulling cloud models is
temporary, once we integrate properly into `/api/tags` we won't need
this anymore.

* Fix server alias syncing

* Update cmd/cmd.go

Co-authored-by: Parth Sareen <parth.sareen@ollama.com>

* address comments

* improve some naming

---------

Co-authored-by: ParthSareen <parth.sareen@ollama.com>
2026-03-03 10:46:33 -08:00

305 lines
6.5 KiB
Go

// Package config provides integration configuration for external coding tools
// (Claude Code, Codex, Droid, OpenCode) to use Ollama models.
package config
import (
"context"
"encoding/json"
"errors"
"fmt"
"os"
"path/filepath"
"strings"
"github.com/ollama/ollama/api"
)
type integration struct {
Models []string `json:"models"`
Aliases map[string]string `json:"aliases,omitempty"`
Onboarded bool `json:"onboarded,omitempty"`
}
type config struct {
Integrations map[string]*integration `json:"integrations"`
LastModel string `json:"last_model,omitempty"`
LastSelection string `json:"last_selection,omitempty"` // "run" or integration name
}
func configPath() (string, error) {
home, err := os.UserHomeDir()
if err != nil {
return "", err
}
return filepath.Join(home, ".ollama", "config.json"), nil
}
func legacyConfigPath() (string, error) {
home, err := os.UserHomeDir()
if err != nil {
return "", err
}
return filepath.Join(home, ".ollama", "config", "config.json"), nil
}
// migrateConfig moves the config from the legacy path to ~/.ollama/config.json
func migrateConfig() (bool, error) {
oldPath, err := legacyConfigPath()
if err != nil {
return false, err
}
oldData, err := os.ReadFile(oldPath)
if err != nil {
if os.IsNotExist(err) {
return false, nil
}
return false, err
}
// Ignore legacy files with invalid JSON and continue startup.
if !json.Valid(oldData) {
return false, nil
}
newPath, err := configPath()
if err != nil {
return false, err
}
if err := os.MkdirAll(filepath.Dir(newPath), 0o755); err != nil {
return false, err
}
if err := os.WriteFile(newPath, oldData, 0o644); err != nil {
return false, fmt.Errorf("write new config: %w", err)
}
_ = os.Remove(oldPath)
_ = os.Remove(filepath.Dir(oldPath)) // clean up empty directory
return true, nil
}
func load() (*config, error) {
path, err := configPath()
if err != nil {
return nil, err
}
data, err := os.ReadFile(path)
if err != nil && os.IsNotExist(err) {
if migrated, merr := migrateConfig(); merr == nil && migrated {
data, err = os.ReadFile(path)
}
}
if err != nil {
if os.IsNotExist(err) {
return &config{Integrations: make(map[string]*integration)}, nil
}
return nil, err
}
var cfg config
if err := json.Unmarshal(data, &cfg); err != nil {
return nil, fmt.Errorf("failed to parse config: %w, at: %s", err, path)
}
if cfg.Integrations == nil {
cfg.Integrations = make(map[string]*integration)
}
return &cfg, nil
}
func save(cfg *config) error {
path, err := configPath()
if err != nil {
return err
}
if err := os.MkdirAll(filepath.Dir(path), 0o755); err != nil {
return err
}
data, err := json.MarshalIndent(cfg, "", " ")
if err != nil {
return err
}
return writeWithBackup(path, data)
}
func SaveIntegration(appName string, models []string) error {
if appName == "" {
return errors.New("app name cannot be empty")
}
cfg, err := load()
if err != nil {
return err
}
key := strings.ToLower(appName)
existing := cfg.Integrations[key]
var aliases map[string]string
var onboarded bool
if existing != nil {
aliases = existing.Aliases
onboarded = existing.Onboarded
}
cfg.Integrations[key] = &integration{
Models: models,
Aliases: aliases,
Onboarded: onboarded,
}
return save(cfg)
}
// integrationOnboarded marks an integration as onboarded in ollama's config.
func integrationOnboarded(appName string) error {
cfg, err := load()
if err != nil {
return err
}
key := strings.ToLower(appName)
existing := cfg.Integrations[key]
if existing == nil {
existing = &integration{}
}
existing.Onboarded = true
cfg.Integrations[key] = existing
return save(cfg)
}
// IntegrationModel returns the first configured model for an integration, or empty string if not configured.
func IntegrationModel(appName string) string {
integrationConfig, err := loadIntegration(appName)
if err != nil || len(integrationConfig.Models) == 0 {
return ""
}
return integrationConfig.Models[0]
}
// IntegrationModels returns all configured models for an integration, or nil.
func IntegrationModels(appName string) []string {
integrationConfig, err := loadIntegration(appName)
if err != nil || len(integrationConfig.Models) == 0 {
return nil
}
return integrationConfig.Models
}
// LastModel returns the last model that was run, or empty string if none.
func LastModel() string {
cfg, err := load()
if err != nil {
return ""
}
return cfg.LastModel
}
// SetLastModel saves the last model that was run.
func SetLastModel(model string) error {
cfg, err := load()
if err != nil {
return err
}
cfg.LastModel = model
return save(cfg)
}
// LastSelection returns the last menu selection ("run" or integration name), or empty string if none.
func LastSelection() string {
cfg, err := load()
if err != nil {
return ""
}
return cfg.LastSelection
}
// SetLastSelection saves the last menu selection ("run" or integration name).
func SetLastSelection(selection string) error {
cfg, err := load()
if err != nil {
return err
}
cfg.LastSelection = selection
return save(cfg)
}
// ModelExists checks if a model exists on the Ollama server.
func ModelExists(ctx context.Context, name string) bool {
if name == "" {
return false
}
if isCloudModelName(name) {
return true
}
client, err := api.ClientFromEnvironment()
if err != nil {
return false
}
models, err := client.List(ctx)
if err != nil {
return false
}
for _, m := range models.Models {
if m.Name == name || strings.HasPrefix(m.Name, name+":") {
return true
}
}
return false
}
func loadIntegration(appName string) (*integration, error) {
cfg, err := load()
if err != nil {
return nil, err
}
integrationConfig, ok := cfg.Integrations[strings.ToLower(appName)]
if !ok {
return nil, os.ErrNotExist
}
return integrationConfig, nil
}
func saveAliases(appName string, aliases map[string]string) error {
if appName == "" {
return errors.New("app name cannot be empty")
}
cfg, err := load()
if err != nil {
return err
}
key := strings.ToLower(appName)
existing := cfg.Integrations[key]
if existing == nil {
existing = &integration{}
}
// Replace aliases entirely (not merge) so deletions are persisted
existing.Aliases = aliases
cfg.Integrations[key] = existing
return save(cfg)
}
func listIntegrations() ([]integration, error) {
cfg, err := load()
if err != nil {
return nil, err
}
result := make([]integration, 0, len(cfg.Integrations))
for _, integrationConfig := range cfg.Integrations {
result = append(result, *integrationConfig)
}
return result, nil
}