Remove GitHub Copilot provider support

Remove all GitHub Copilot integration from the codebase to streamline
the supported provider set to OpenAI, OpenRouter, Anthropic, and Ollama.

Changes:
- Delete core Copilot implementation (copilot.go) and all related tests
- Remove Copilot configuration fields from App struct and Config
- Remove Copilot from provider factory and API key handling
- Update all test files to replace Copilot references with other providers
- Remove Copilot documentation from README, configuration guide, and examples
- Remove Copilot section from config.toml.example

All tests pass successfully after removal.

Co-authored-by: Cursor <cursoragent@cursor.com>

23 files changed, 52 insertions, 909 deletions

diff --git a/internal/appconfig/config.go b/internal/appconfig/config.go
index c9af85e..3077d42 100644
--- a/internal/appconfig/config.go
+++ b/internal/appconfig/config.go
@@ -69,12 +69,8 @@ type App struct {
 	OllamaModel           string   `json:"ollama_model" toml:"ollama_model"`
 	// Default temperature for Ollama requests (nil means use provider default)
 	OllamaTemperature *float64 `json:"ollama_temperature" toml:"ollama_temperature"`
-	CopilotBaseURL    string   `json:"copilot_base_url" toml:"copilot_base_url"`
-	CopilotModel      string   `json:"copilot_model" toml:"copilot_model"`
-	// Default temperature for Copilot requests (nil means use provider default)
-	CopilotTemperature *float64 `json:"copilot_temperature" toml:"copilot_temperature"`
-	AnthropicBaseURL   string   `json:"anthropic_base_url" toml:"anthropic_base_url"`
-	AnthropicModel     string   `json:"anthropic_model" toml:"anthropic_model"`
+	AnthropicBaseURL  string   `json:"anthropic_base_url" toml:"anthropic_base_url"`
+	AnthropicModel    string   `json:"anthropic_model" toml:"anthropic_model"`
 	// Default temperature for Anthropic requests (nil means use provider default)
 	AnthropicTemperature *float64 `json:"anthropic_temperature" toml:"anthropic_temperature"`
 
@@ -146,7 +142,6 @@ func newDefaultConfig() App {
 		CodingTemperature:     &t,
 		OpenAITemperature:     &t,
 		OllamaTemperature:     &t,
-		CopilotTemperature:    &t,
 		AnthropicTemperature:  &t,
 		ManualInvokeMinPrefix: 0,
 		CompletionDebounceMs:  800,
@@ -244,7 +239,6 @@ type fileConfig struct {
 	Provider   sectionProvider   `toml:"provider"`
 	OpenAI     sectionOpenAI     `toml:"openai"`
 	OpenRouter sectionOpenRouter `toml:"openrouter"`
-	Copilot    sectionCopilot    `toml:"copilot"`
 	Ollama     sectionOllama     `toml:"ollama"`
 	Anthropic  sectionAnthropic  `toml:"anthropic"`
 	Prompts    sectionPrompts    `toml:"prompts"`
@@ -333,12 +327,6 @@ type sectionOpenRouter struct {
 	Temperature *float64 `toml:"temperature"`
 }
 
-type sectionCopilot struct {
-	Model       string   `toml:"model"`
-	BaseURL     string   `toml:"base_url"`
-	Temperature *float64 `toml:"temperature"`
-}
-
 type sectionOllama struct {
 	Model       string   `toml:"model"`
 	BaseURL     string   `toml:"base_url"`
@@ -489,16 +477,6 @@ func (fc *fileConfig) toApp() App {
 		out.mergeProviderFields(&tmp)
 	}
 
-	// copilot
-	if (fc.Copilot != sectionCopilot{}) || fc.Copilot.Temperature != nil {
-		tmp := App{
-			CopilotBaseURL:     fc.Copilot.BaseURL,
-			CopilotModel:       fc.Copilot.Model,
-			CopilotTemperature: fc.Copilot.Temperature,
-		}
-		out.mergeProviderFields(&tmp)
-	}
-
 	// ollama
 	if (fc.Ollama != sectionOllama{}) || fc.Ollama.Temperature != nil {
 		tmp := App{
@@ -658,10 +636,9 @@ func loadFromFile(path string, logger *log.Logger) (*App, error) {
 		"chat_suffix": {}, "chat_prefixes": {}, "coding_temperature": {}, "provider": {},
 		"openai_model": {}, "openai_base_url": {}, "openai_temperature": {},
 		"ollama_model": {}, "ollama_base_url": {}, "ollama_temperature": {},
-		"copilot_model": {}, "copilot_base_url": {}, "copilot_temperature": {},
 	}
 	for k := range raw {
-		if _, isTable := map[string]struct{}{"general": {}, "logging": {}, "completion": {}, "triggers": {}, "inline": {}, "chat": {}, "provider": {}, "models": {}, "openai": {}, "copilot": {}, "ollama": {}, "prompts": {}}[k]; isTable {
+		if _, isTable := map[string]struct{}{"general": {}, "logging": {}, "completion": {}, "triggers": {}, "inline": {}, "chat": {}, "provider": {}, "models": {}, "openai": {}, "ollama": {}, "prompts": {}}[k]; isTable {
 			continue
 		}
 		if _, isLegacy := legacy[k]; isLegacy {
@@ -1103,15 +1080,6 @@ func (a *App) mergeProviderFields(other *App) {
 	if other.OllamaTemperature != nil { // allow explicit 0.0
 		a.OllamaTemperature = other.OllamaTemperature
 	}
-	if s := strings.TrimSpace(other.CopilotBaseURL); s != "" {
-		a.CopilotBaseURL = s
-	}
-	if s := strings.TrimSpace(other.CopilotModel); s != "" {
-		a.CopilotModel = s
-	}
-	if other.CopilotTemperature != nil { // allow explicit 0.0
-		a.CopilotTemperature = other.CopilotTemperature
-	}
 	if s := strings.TrimSpace(other.AnthropicBaseURL); s != "" {
 		a.AnthropicBaseURL = s
 	}
@@ -1331,19 +1299,6 @@ func loadFromEnv(logger *log.Logger) *App {
 		any = true
 	}
 
-	if s := getenv("HEXAI_COPILOT_BASE_URL"); s != "" {
-		out.CopilotBaseURL = s
-		any = true
-	}
-	if model, ok := pickModel("copilot", getenv("HEXAI_COPILOT_MODEL")); ok {
-		out.CopilotModel = model
-		any = true
-	}
-	if f, ok := parseFloatPtr("HEXAI_COPILOT_TEMPERATURE"); ok {
-		out.CopilotTemperature = f
-		any = true
-	}
-
 	if s := getenv("HEXAI_ANTHROPIC_BASE_URL"); s != "" {
 		out.AnthropicBaseURL = s
 		any = true
diff --git a/internal/appconfig/config_env_model_test.go b/internal/appconfig/config_env_model_test.go
index 7038819..e10fa5d 100644
--- a/internal/appconfig/config_env_model_test.go
+++ b/internal/appconfig/config_env_model_test.go
@@ -37,9 +37,9 @@ func TestEnv_ModelForce_OverridesProviderSpecific(t *testing.T) {
 }
 
 func TestEnv_SurfaceModelOverrides(t *testing.T) {
-	t.Setenv("HEXAI_MODEL_COMPLETION", "gpt-c")
+	t.Setenv("HEXAI_MODEL_COMPLETION", "claude-c")
 	t.Setenv("HEXAI_TEMPERATURE_COMPLETION", "0.44")
-	t.Setenv("HEXAI_PROVIDER_COMPLETION", "copilot")
+	t.Setenv("HEXAI_PROVIDER_COMPLETION", "anthropic")
 	t.Setenv("HEXAI_MODEL_CLI", "gpt-cli")
 	t.Setenv("HEXAI_TEMPERATURE_CLI", "0.22")
 	t.Setenv("HEXAI_PROVIDER_CLI", "ollama")
@@ -48,13 +48,13 @@ func TestEnv_SurfaceModelOverrides(t *testing.T) {
 		t.Fatalf("expected single completion entry, got %+v", cfg.CompletionConfigs)
 	}
 	comp := cfg.CompletionConfigs[0]
-	if comp.Model != "gpt-c" {
+	if comp.Model != "claude-c" {
 		t.Fatalf("expected completion model override, got %+v", comp)
 	}
 	if comp.Temperature == nil || *comp.Temperature != 0.44 {
 		t.Fatalf("expected completion temperature override, got %+v", comp)
 	}
-	if comp.Provider != "copilot" {
+	if comp.Provider != "anthropic" {
 		t.Fatalf("expected completion provider override, got %+v", comp)
 	}
 	if len(cfg.CLIConfigs) != 1 {
diff --git a/internal/appconfig/config_test.go b/internal/appconfig/config_test.go
index 2c00f68..ff9616b 100644
--- a/internal/appconfig/config_test.go
+++ b/internal/appconfig/config_test.go
@@ -75,8 +75,8 @@ func TestParseSurfaceModels_CodeActionWarns(t *testing.T) {
   model = "gpt-4o"
 
   [[models.code_action]]
-  provider = "copilot"
-  model = "cpt"
+  provider = "anthropic"
+  model = "claude"
 `)
 	var buf bytes.Buffer
 	logger := log.New(&buf, "", 0)
@@ -121,9 +121,9 @@ model = "gpt-file-complete"
 provider = "openai"
 
 [[models.code_action]]
-model = "gpt-file-action"
+model = "claude-file-action"
 temperature = 0.45
-provider = "copilot"
+provider = "anthropic"
 
 [[models.chat]]
 model = "gpt-file-chat"
@@ -146,11 +146,6 @@ temperature = 0.0
 base_url = "http://ollama"
 model = "llama"
 temperature = 0.0
-
-[copilot]
-base_url = "http://copilot"
-model = "ghost"
-temperature = 0.0
 `)
 
 	if _, err := loadFromFile(cfgPath, newLogger()); err != nil {
@@ -175,18 +170,15 @@ temperature = 0.0
 	withEnv(t, "HEXAI_OLLAMA_BASE_URL", "http://ollama-override")
 	withEnv(t, "HEXAI_OLLAMA_MODEL", "mistral")
 	withEnv(t, "HEXAI_OLLAMA_TEMPERATURE", "0.6")
-	withEnv(t, "HEXAI_COPILOT_BASE_URL", "http://copilot-override")
-	withEnv(t, "HEXAI_COPILOT_MODEL", "ghost-override")
-	withEnv(t, "HEXAI_COPILOT_TEMPERATURE", "0.3")
 	withEnv(t, "HEXAI_MODEL_COMPLETION", "env-completion")
 	withEnv(t, "HEXAI_TEMPERATURE_COMPLETION", "0.33")
-	withEnv(t, "HEXAI_PROVIDER_COMPLETION", "copilot")
+	withEnv(t, "HEXAI_PROVIDER_COMPLETION", "ollama")
 	withEnv(t, "HEXAI_MODEL_CODE_ACTION", "env-action")
 	withEnv(t, "HEXAI_TEMPERATURE_CODE_ACTION", "0.55")
 	withEnv(t, "HEXAI_PROVIDER_CODE_ACTION", "openai")
 	withEnv(t, "HEXAI_MODEL_CHAT", "env-chat")
 	withEnv(t, "HEXAI_TEMPERATURE_CHAT", "0.66")
-	withEnv(t, "HEXAI_PROVIDER_CHAT", "copilot")
+	withEnv(t, "HEXAI_PROVIDER_CHAT", "anthropic")
 	withEnv(t, "HEXAI_MODEL_CLI", "env-cli")
 	withEnv(t, "HEXAI_TEMPERATURE_CLI", "0.77")
 	withEnv(t, "HEXAI_PROVIDER_CLI", "ollama")
@@ -217,16 +209,13 @@ temperature = 0.0
 	if cfg.OllamaBaseURL != "http://ollama-override" || cfg.OllamaModel != "mistral" || cfg.OllamaTemperature == nil || *cfg.OllamaTemperature != 0.6 {
 		t.Fatalf("ollama overrides not applied: %+v", cfg)
 	}
-	if cfg.CopilotBaseURL != "http://copilot-override" || cfg.CopilotModel != "ghost-override" || cfg.CopilotTemperature == nil || *cfg.CopilotTemperature != 0.3 {
-		t.Fatalf("copilot overrides not applied: %+v", cfg)
-	}
 	if len(cfg.CompletionConfigs) != 1 || cfg.CompletionConfigs[0].Model != "env-completion" {
 		t.Fatalf("completion overrides not applied: %+v", cfg.CompletionConfigs)
 	}
 	if cfg.CompletionConfigs[0].Temperature == nil || *cfg.CompletionConfigs[0].Temperature != 0.33 {
 		t.Fatalf("completion temperature override missing: %+v", cfg.CompletionConfigs[0])
 	}
-	if cfg.CompletionConfigs[0].Provider != "copilot" {
+	if cfg.CompletionConfigs[0].Provider != "ollama" {
 		t.Fatalf("completion provider override not applied: %+v", cfg.CompletionConfigs[0])
 	}
 	if len(cfg.CodeActionConfigs) != 1 || cfg.CodeActionConfigs[0].Model != "env-action" {
@@ -244,7 +233,7 @@ temperature = 0.0
 	if cfg.ChatConfigs[0].Temperature == nil || *cfg.ChatConfigs[0].Temperature != 0.66 {
 		t.Fatalf("chat temp override missing: %+v", cfg.ChatConfigs[0])
 	}
-	if cfg.ChatConfigs[0].Provider != "copilot" {
+	if cfg.ChatConfigs[0].Provider != "anthropic" {
 		t.Fatalf("chat provider override not applied: %+v", cfg.ChatConfigs[0])
 	}
 	if len(cfg.CLIConfigs) != 1 || cfg.CLIConfigs[0].Model != "env-cli" {
@@ -260,7 +249,7 @@ temperature = 0.0
 	// Ensure file values would have applied absent env
 	// Spot-check: reset env and reload
 	for _, k := range []string{
-		"HEXAI_MAX_TOKENS", "HEXAI_CONTEXT_MODE", "HEXAI_CONTEXT_WINDOW_LINES", "HEXAI_MAX_CONTEXT_TOKENS", "HEXAI_LOG_PREVIEW_LIMIT", "HEXAI_CODING_TEMPERATURE", "HEXAI_MANUAL_INVOKE_MIN_PREFIX", "HEXAI_COMPLETION_DEBOUNCE_MS", "HEXAI_COMPLETION_THROTTLE_MS", "HEXAI_TRIGGER_CHARACTERS", "HEXAI_PROVIDER", "HEXAI_OPENAI_BASE_URL", "HEXAI_OPENAI_MODEL", "HEXAI_OPENAI_TEMPERATURE", "HEXAI_OLLAMA_BASE_URL", "HEXAI_OLLAMA_MODEL", "HEXAI_OLLAMA_TEMPERATURE", "HEXAI_COPILOT_BASE_URL", "HEXAI_COPILOT_MODEL", "HEXAI_COPILOT_TEMPERATURE", "HEXAI_MODEL_COMPLETION", "HEXAI_TEMPERATURE_COMPLETION", "HEXAI_MODEL_CODE_ACTION", "HEXAI_TEMPERATURE_CODE_ACTION", "HEXAI_MODEL_CHAT", "HEXAI_TEMPERATURE_CHAT", "HEXAI_MODEL_CLI", "HEXAI_TEMPERATURE_CLI", "HEXAI_PROVIDER_COMPLETION", "HEXAI_PROVIDER_CODE_ACTION", "HEXAI_PROVIDER_CHAT", "HEXAI_PROVIDER_CLI",
+		"HEXAI_MAX_TOKENS", "HEXAI_CONTEXT_MODE", "HEXAI_CONTEXT_WINDOW_LINES", "HEXAI_MAX_CONTEXT_TOKENS", "HEXAI_LOG_PREVIEW_LIMIT", "HEXAI_CODING_TEMPERATURE", "HEXAI_MANUAL_INVOKE_MIN_PREFIX", "HEXAI_COMPLETION_DEBOUNCE_MS", "HEXAI_COMPLETION_THROTTLE_MS", "HEXAI_TRIGGER_CHARACTERS", "HEXAI_PROVIDER", "HEXAI_OPENAI_BASE_URL", "HEXAI_OPENAI_MODEL", "HEXAI_OPENAI_TEMPERATURE", "HEXAI_OLLAMA_BASE_URL", "HEXAI_OLLAMA_MODEL", "HEXAI_OLLAMA_TEMPERATURE", "HEXAI_MODEL_COMPLETION", "HEXAI_TEMPERATURE_COMPLETION", "HEXAI_MODEL_CODE_ACTION", "HEXAI_TEMPERATURE_CODE_ACTION", "HEXAI_MODEL_CHAT", "HEXAI_TEMPERATURE_CHAT", "HEXAI_MODEL_CLI", "HEXAI_TEMPERATURE_CLI", "HEXAI_PROVIDER_COMPLETION", "HEXAI_PROVIDER_CODE_ACTION", "HEXAI_PROVIDER_CHAT", "HEXAI_PROVIDER_CLI",
 	} {
 		t.Setenv(k, "")
 	}
@@ -283,13 +272,13 @@ temperature = 0.0
 	if cfg2.CompletionConfigs[0].Provider != "openai" {
 		t.Fatalf("file merge (completion provider) not applied: %+v", cfg2.CompletionConfigs[0])
 	}
-	if len(cfg2.CodeActionConfigs) != 1 || cfg2.CodeActionConfigs[0].Model != "gpt-file-action" {
+	if len(cfg2.CodeActionConfigs) != 1 || cfg2.CodeActionConfigs[0].Model != "claude-file-action" {
 		t.Fatalf("file merge (code action) not applied: %+v", cfg2.CodeActionConfigs)
 	}
 	if cfg2.CodeActionConfigs[0].Temperature == nil || *cfg2.CodeActionConfigs[0].Temperature != 0.45 {
 		t.Fatalf("expected code action temp 0.45, got %+v", cfg2.CodeActionConfigs[0])
 	}
-	if cfg2.CodeActionConfigs[0].Provider != "copilot" {
+	if cfg2.CodeActionConfigs[0].Provider != "anthropic" {
 		t.Fatalf("file merge (code action provider) not applied: %+v", cfg2.CodeActionConfigs[0])
 	}
 	if len(cfg2.ChatConfigs) != 1 || cfg2.ChatConfigs[0].Model != "gpt-file-chat" {
@@ -384,11 +373,6 @@ temperature = 0.0
 model = "mistral"
 base_url = "http://ollama"
 temperature = 0.0
-
-[copilot]
-model = "ghost"
-base_url = "http://copilot"
-temperature = 0.0
 `
 	writeFile(t, cfgPath, content)
 
@@ -418,9 +402,6 @@ temperature = 0.0
 	if cfg.OllamaModel != "mistral" || cfg.OllamaBaseURL != "http://ollama" || cfg.OllamaTemperature == nil || *cfg.OllamaTemperature != 0.0 {
 		t.Fatalf("sectioned ollama wrong: %+v", cfg)
 	}
-	if cfg.CopilotModel != "ghost" || cfg.CopilotBaseURL != "http://copilot" || cfg.CopilotTemperature == nil || *cfg.CopilotTemperature != 0.0 {
-		t.Fatalf("sectioned copilot wrong: %+v", cfg)
-	}
 }
 
 func TestLoad_FileTables_Prompts_AllSections(t *testing.T) {
diff --git a/internal/hexaiaction/prompts.go b/internal/hexaiaction/prompts.go
index a113391..fc743a0 100644
--- a/internal/hexaiaction/prompts.go
+++ b/internal/hexaiaction/prompts.go
@@ -49,8 +49,8 @@ func defaultModelForProvider(cfg appconfig.App, provider string) string {
 	switch provider {
 	case "ollama":
 		return cfg.OllamaModel
-	case "copilot":
-		return cfg.CopilotModel
+	case "anthropic":
+		return cfg.AnthropicModel
 	default:
 		return cfg.OpenAIModel
 	}
diff --git a/internal/hexaiaction/prompts_more_test.go b/internal/hexaiaction/prompts_more_test.go
index cfccd0c..a4410e5 100644
--- a/internal/hexaiaction/prompts_more_test.go
+++ b/internal/hexaiaction/prompts_more_test.go
@@ -35,8 +35,8 @@ func TestReqOptsFrom_Override(t *testing.T) {
 	cfg := appconfig.App{
 		MaxTokens:         123,
 		Provider:          "openai",
-		CopilotModel:      "gpt-4o",
-		CodeActionConfigs: []appconfig.SurfaceConfig{{Provider: "copilot", Model: "override", Temperature: ptrFloat(0.6)}},
+		AnthropicModel:    "claude-3-5-sonnet",
+		CodeActionConfigs: []appconfig.SurfaceConfig{{Provider: "anthropic", Model: "override", Temperature: ptrFloat(0.6)}},
 	}
 	req := reqOptsFrom(cfg)
 	if req.model != "override" {
diff --git a/internal/hexaiaction/run_more_test.go b/internal/hexaiaction/run_more_test.go
index a3e7f25..57bd933 100644
--- a/internal/hexaiaction/run_more_test.go
+++ b/internal/hexaiaction/run_more_test.go
@@ -14,7 +14,7 @@ import (
 // Covers the early error path in Run when no API key is available for the default provider.
 func TestRun_MissingAPIKey(t *testing.T) {
 	// Ensure no provider API keys in env
-	for _, k := range []string{"HEXAI_OPENAI_API_KEY", "OPENAI_API_KEY", "HEXAI_COPILOT_API_KEY", "COPILOT_API_KEY"} {
+	for _, k := range []string{"HEXAI_OPENAI_API_KEY", "OPENAI_API_KEY"} {
 		t.Setenv(k, "")
 	}
 	// Provide minimal stdin to get past empty input check (if reached)
diff --git a/internal/hexaicli/run.go b/internal/hexaicli/run.go
index 7b360e9..9ea3a40 100644
--- a/internal/hexaicli/run.go
+++ b/internal/hexaicli/run.go
@@ -77,10 +77,6 @@ func buildCLIJobs(cfg appconfig.App) ([]cliJob, error) {
 			if entry.Model != "" {
 				derived.OpenAIModel = entry.Model
 			}
-		case "copilot":
-			if entry.Model != "" {
-				derived.CopilotModel = entry.Model
-			}
 		case "ollama":
 			if entry.Model != "" {
 				derived.OllamaModel = entry.Model
@@ -151,8 +147,8 @@ func defaultModelForProvider(cfg appconfig.App, provider string) string {
 	switch provider {
 	case "ollama":
 		return cfg.OllamaModel
-	case "copilot":
-		return cfg.CopilotModel
+	case "anthropic":
+		return cfg.AnthropicModel
 	default:
 		return cfg.OpenAIModel
 	}
diff --git a/internal/hexaicli/run_test.go b/internal/hexaicli/run_test.go
index 43576cf..34a5c51 100644
--- a/internal/hexaicli/run_test.go
+++ b/internal/hexaicli/run_test.go
@@ -161,11 +161,11 @@ func TestPrintProviderInfo(t *testing.T) {
 
 func TestBuildCLIRequest_Override(t *testing.T) {
 	cfg := appconfig.App{
-		Provider:     "openai",
-		CopilotModel: "gpt-4o",
+		Provider:       "openai",
+		AnthropicModel: "claude-3-5-sonnet",
 	}
-	entry := appconfig.SurfaceConfig{Provider: "copilot", Model: "override", Temperature: floatPtr(0.7)}
-	req := buildCLIRequest(entry, "copilot", cfg, &fakeClient{name: "copilot", model: "default"})
+	entry := appconfig.SurfaceConfig{Provider: "anthropic", Model: "override", Temperature: floatPtr(0.7)}
+	req := buildCLIRequest(entry, "anthropic", cfg, &fakeClient{name: "anthropic", model: "default"})
 	if req.model != "override" {
 		t.Fatalf("expected model override, got %q", req.model)
 	}
@@ -199,8 +199,8 @@ func TestBuildCLIJobs_MultiEntries(t *testing.T) {
 	defer func() { newClientFromApp = old }()
 	newClientFromApp = func(cfg appconfig.App) (llm.Client, error) {
 		model := cfg.OpenAIModel
-		if cfg.Provider == "copilot" {
-			model = cfg.CopilotModel
+		if cfg.Provider == "anthropic" {
+			model = cfg.AnthropicModel
 		}
 		if cfg.Provider == "ollama" {
 			model = cfg.OllamaModel
@@ -215,7 +215,7 @@ func TestBuildCLIJobs_MultiEntries(t *testing.T) {
 		OllamaModel: "llama3",
 		CLIConfigs: []appconfig.SurfaceConfig{
 			{Provider: "openai", Model: "gpt-4o"},
-			{Provider: "copilot", Model: "cpt"},
+			{Provider: "anthropic", Model: "claude"},
 		},
 	}
 	jobs, err := buildCLIJobs(cfg)
@@ -228,18 +228,18 @@ func TestBuildCLIJobs_MultiEntries(t *testing.T) {
 	if jobs[0].provider != "openai" || jobs[0].req.model != "gpt-4o" {
 		t.Fatalf("unexpected first job: %+v", jobs[0])
 	}
-	if jobs[1].provider != "copilot" || jobs[1].req.model != "cpt" {
+	if jobs[1].provider != "anthropic" || jobs[1].req.model != "claude" {
 		t.Fatalf("unexpected second job: %+v", jobs[1])
 	}
 }
 
 func TestFilterJobsBySelection(t *testing.T) {
-	jobs := []cliJob{{index: 0, provider: "openai"}, {index: 1, provider: "ollama"}, {index: 2, provider: "copilot"}}
+	jobs := []cliJob{{index: 0, provider: "openai"}, {index: 1, provider: "ollama"}, {index: 2, provider: "anthropic"}}
 	filtered, err := filterJobsBySelection(jobs, []int{2, 0})
 	if err != nil {
 		t.Fatalf("unexpected error: %v", err)
 	}
-	if len(filtered) != 2 || filtered[0].provider != "copilot" || filtered[1].provider != "openai" {
+	if len(filtered) != 2 || filtered[0].provider != "anthropic" || filtered[1].provider != "openai" {
 		t.Fatalf("unexpected filtered order: %+v", filtered)
 	}
 	if filtered[0].index != 0 || filtered[1].index != 1 {
diff --git a/internal/hexaicli/testhelpers_test.go b/internal/hexaicli/testhelpers_test.go
index 8f6863d..3197880 100644
--- a/internal/hexaicli/testhelpers_test.go
+++ b/internal/hexaicli/testhelpers_test.go
@@ -63,7 +63,6 @@ func (s *fakeStreamer) ChatStream(ctx context.Context, messages []llm.Message, o
 	return nil
 }
 
-
 func writeConfigString(t *testing.T, path string, contents string) {
 	t.Helper()
 	if err := os.MkdirAll(filepath.Dir(path), 0o755); err != nil {
diff --git a/internal/hexailsp/run.go b/internal/hexailsp/run.go
index 47ed648..e2aaf9d 100644
--- a/internal/hexailsp/run.go
+++ b/internal/hexailsp/run.go
@@ -119,9 +119,6 @@ func buildClientIfNil(cfg appconfig.App, client llm.Client) llm.Client {
 		OllamaBaseURL:         cfg.OllamaBaseURL,
 		OllamaModel:           cfg.OllamaModel,
 		OllamaTemperature:     cfg.OllamaTemperature,
-		CopilotBaseURL:        cfg.CopilotBaseURL,
-		CopilotModel:          cfg.CopilotModel,
-		CopilotTemperature:    cfg.CopilotTemperature,
 		AnthropicBaseURL:      cfg.AnthropicBaseURL,
 		AnthropicModel:        cfg.AnthropicModel,
 		AnthropicTemperature:  cfg.AnthropicTemperature,
@@ -136,17 +133,12 @@ func buildClientIfNil(cfg appconfig.App, client llm.Client) llm.Client {
 	if strings.TrimSpace(orKey) == "" {
 		orKey = os.Getenv("OPENROUTER_API_KEY")
 	}
-	// Prefer HEXAI_COPILOT_API_KEY; fall back to COPILOT_API_KEY
-	cpKey := os.Getenv("HEXAI_COPILOT_API_KEY")
-	if strings.TrimSpace(cpKey) == "" {
-		cpKey = os.Getenv("COPILOT_API_KEY")
-	}
 	// Prefer HEXAI_ANTHROPIC_API_KEY; fall back to ANTHROPIC_API_KEY
 	anKey := os.Getenv("HEXAI_ANTHROPIC_API_KEY")
 	if strings.TrimSpace(anKey) == "" {
 		anKey = os.Getenv("ANTHROPIC_API_KEY")
 	}
-	if c, err := llm.NewFromConfig(llmCfg, oaKey, orKey, cpKey, anKey); err != nil {
+	if c, err := llm.NewFromConfig(llmCfg, oaKey, orKey, anKey); err != nil {
 		logging.Logf("lsp ", "llm disabled: %v", err)
 		return nil
 	} else {
diff --git a/internal/llm/copilot.go b/internal/llm/copilot.go
deleted file mode 100644
index 43419ea..0000000
--- a/internal/llm/copilot.go
+++ /dev/null
@@ -1,412 +0,0 @@
-// Summary: GitHub Copilot client for chat and Codex-style code completion.
-package llm
-
-import (
-	"bytes"
-	"context"
-	"encoding/base64"
-	"encoding/json"
-	"errors"
-	"fmt"
-	"io"
-	"net/http"
-	"regexp"
-	"strings"
-	"time"
-
-	appver "codeberg.org/snonux/hexai/internal"
-	"codeberg.org/snonux/hexai/internal/logging"
-)
-
-// copilotClient implements Client against GitHub Copilot's Chat Completions API.
-type copilotClient struct {
-	httpClient         *http.Client
-	apiKey             string
-	baseURL            string
-	defaultModel       string
-	chatLogger         logging.ChatLogger
-	defaultTemperature *float64
-
-	// cached Copilot session token retrieved from GitHub API using apiKey
-	sessionToken string
-	tokenExpiry  time.Time
-}
-
-type copilotChatRequest struct {
-	Model       string           `json:"model"`
-	Messages    []copilotMessage `json:"messages"`
-	Temperature *float64         `json:"temperature,omitempty"`
-	MaxTokens   *int             `json:"max_tokens,omitempty"`
-	Stop        []string         `json:"stop,omitempty"`
-}
-
-type copilotMessage struct {
-	Role    string `json:"role"`
-	Content string `json:"content"`
-}
-
-type copilotChatResponse struct {
-	Choices []struct {
-		Index   int `json:"index"`
-		Message struct {
-			Role    string `json:"role"`
-			Content string `json:"content"`
-		} `json:"message"`
-		FinishReason string `json:"finish_reason"`
-	} `json:"choices"`
-	Error *struct {
-		Message string `json:"message"`
-		Type    string `json:"type"`
-		Param   any    `json:"param"`
-		Code    any    `json:"code"`
-	} `json:"error,omitempty"`
-}
-
-// Constructor (kept among the first functions by convention)
-func newCopilot(baseURL, model, apiKey string, defaultTemp *float64) Client {
-	return newCopilotWithTimeout(baseURL, model, apiKey, defaultTemp, 0)
-}
-
-func newCopilotWithTimeout(baseURL, model, apiKey string, defaultTemp *float64, timeoutSec int) Client {
-	if strings.TrimSpace(baseURL) == "" {
-		baseURL = "https://api.githubcopilot.com"
-	}
-	if strings.TrimSpace(model) == "" {
-		// GitHub Models (Copilot API) commonly supports gpt-4o/gpt-4o-mini.
-		// Default to a broadly available, cost-effective option.
-		model = "gpt-4o-mini"
-	}
-	if timeoutSec <= 0 {
-		timeoutSec = 30
-	}
-	return copilotClient{
-		httpClient:         &http.Client{Timeout: time.Duration(timeoutSec) * time.Second},
-		apiKey:             apiKey,
-		baseURL:            strings.TrimRight(baseURL, "/"),
-		defaultModel:       model,
-		chatLogger:         logging.NewChatLogger("copilot"),
-		defaultTemperature: defaultTemp,
-	}
-}
-
-func (c copilotClient) Chat(ctx context.Context, messages []Message, opts ...RequestOption) (string, error) {
-	if strings.TrimSpace(c.apiKey) == "" {
-		return nilStringErr("missing Copilot API key")
-	}
-	// Ensure we have a fresh session token
-	if err := c.ensureSession(ctx); err != nil {
-		return "", err
-	}
-	o := Options{Model: c.defaultModel}
-	for _, opt := range opts {
-		opt(&o)
-	}
-	if o.Model == "" {
-		o.Model = c.defaultModel
-	}
-	start := time.Now()
-	logMessages := make([]struct{ Role, Content string }, len(messages))
-	for i, m := range messages {
-		logMessages[i] = struct{ Role, Content string }{m.Role, m.Content}
-	}
-	c.chatLogger.LogStart(false, o.Model, o.Temperature, o.MaxTokens, o.Stop, logMessages)
-
-	req := buildCopilotChatRequest(o, messages, c.defaultTemperature)
-	body, err := json.Marshal(req)
-	if err != nil {
-		logging.Logf("llm/copilot ", "marshal error: %v", err)
-		return "", err
-	}
-
-	endpoint := c.baseURL + "/chat/completions"
-	logging.Logf("llm/copilot ", "POST %s", endpoint)
-	resp, err := c.postJSON(ctx, endpoint, body, c.headersChat())
-	if err != nil {
-		logging.Logf("llm/copilot ", "%shttp error after %s: %v%s", logging.AnsiRed, time.Since(start), err, logging.AnsiBase)
-		return "", err
-	}
-	defer func() {
-		if err := resp.Body.Close(); err != nil {
-			logging.Logf("llm/copilot", "failed to close response body: %v", err)
-		}
-	}()
-	if err := handleCopilotNon2xx(resp, start); err != nil {
-		return "", err
-	}
-	out, err := decodeCopilotChat(resp, start)
-	if err != nil {
-		return "", err
-	}
-	if len(out.Choices) == 0 {
-		logging.Logf("llm/copilot ", "%sno choices returned duration=%s%s", logging.AnsiRed, time.Since(start), logging.AnsiBase)
-		return "", errors.New("copilot: no choices returned")
-	}
-	content := out.Choices[0].Message.Content
-	logging.Logf("llm/copilot ", "success choice=0 finish=%s size=%d preview=%s%s%s duration=%s", out.Choices[0].FinishReason, len(content), logging.AnsiGreen, logging.PreviewForLog(content), logging.AnsiBase, time.Since(start))
-	return content, nil
-}
-
-// Provider metadata
-func (c copilotClient) Name() string         { return "copilot" }
-func (c copilotClient) DefaultModel() string { return c.defaultModel }
-
-// helpers
-func buildCopilotChatRequest(o Options, messages []Message, defaultTemp *float64) copilotChatRequest {
-	req := copilotChatRequest{Model: o.Model}
-	req.Messages = make([]copilotMessage, len(messages))
-	for i, m := range messages {
-		req.Messages[i] = copilotMessage(m)
-	}
-	if o.Temperature != 0 {
-		req.Temperature = &o.Temperature
-	} else if defaultTemp != nil {
-		t := *defaultTemp
-		req.Temperature = &t
-	}
-	if o.MaxTokens > 0 {
-		req.MaxTokens = &o.MaxTokens
-	}
-	if len(o.Stop) > 0 {
-		req.Stop = o.Stop
-	}
-	return req
-}
-
-func (c copilotClient) postJSON(ctx context.Context, url string, body []byte, headers map[string]string) (*http.Response, error) {
-	req, err := http.NewRequestWithContext(ctx, http.MethodPost, url, bytes.NewReader(body))
-	if err != nil {
-		return nil, err
-	}
-	for k, v := range headers {
-		req.Header.Set(k, v)
-	}
-	return c.httpClient.Do(req)
-}
-
-func handleCopilotNon2xx(resp *http.Response, start time.Time) error {
-	if resp.StatusCode >= 200 && resp.StatusCode < 300 {
-		return nil
-	}
-	var apiErr copilotChatResponse
-	_ = json.NewDecoder(resp.Body).Decode(&apiErr)
-	if apiErr.Error != nil && strings.TrimSpace(apiErr.Error.Message) != "" {
-		logging.Logf("llm/copilot ", "%sapi error status=%d type=%s msg=%s duration=%s%s", logging.AnsiRed, resp.StatusCode, apiErr.Error.Type, apiErr.Error.Message, time.Since(start), logging.AnsiBase)
-		return fmt.Errorf("copilot error: %s (status %d)", apiErr.Error.Message, resp.StatusCode)
-	}
-	logging.Logf("llm/copilot ", "%shttp non-2xx status=%d duration=%s%s", logging.AnsiRed, resp.StatusCode, time.Since(start), logging.AnsiBase)
-	return fmt.Errorf("copilot http error: status %d", resp.StatusCode)
-}
-
-func decodeCopilotChat(resp *http.Response, start time.Time) (copilotChatResponse, error) {
-	var out copilotChatResponse
-	if err := json.NewDecoder(resp.Body).Decode(&out); err != nil {
-		logging.Logf("llm/copilot ", "%sdecode error after %s: %v%s", logging.AnsiRed, time.Since(start), err, logging.AnsiBase)
-		return copilotChatResponse{}, err
-	}
-	return out, nil
-}
-
-// --- Copilot session token management ---
-
-type ghCopilotTokenResp struct {
-	Token string `json:"token"`
-}
-
-func (c *copilotClient) ensureSession(ctx context.Context) error {
-	// If token valid for >60s, reuse
-	if c.sessionToken != "" && time.Now().Add(60*time.Second).Before(c.tokenExpiry) {
-		return nil
-	}
-	if strings.TrimSpace(c.apiKey) == "" {
-		return errors.New("missing Copilot API key")
-	}
-	req, err := http.NewRequestWithContext(ctx, http.MethodGet, "https://api.github.com/copilot_internal/v2/token", nil)
-	if err != nil {
-		return err
-	}
-	req.Header.Set("Authorization", "Bearer "+c.apiKey)
-	req.Header.Set("Accept", "application/json")
-	req.Header.Set("User-Agent", "hexai/"+appver.Version)
-	resp, err := c.httpClient.Do(req)
-	if err != nil {
-		return err
-	}
-	defer func() {
-		if err := resp.Body.Close(); err != nil {
-			logging.Logf("llm/copilot", "failed to close response body: %v", err)
-		}
-	}()
-	if resp.StatusCode < 200 || resp.StatusCode >= 300 {
-		return fmt.Errorf("copilot token http error: %d", resp.StatusCode)
-	}
-	var out ghCopilotTokenResp
-	if err := json.NewDecoder(resp.Body).Decode(&out); err != nil {
-		return err
-	}
-	if strings.TrimSpace(out.Token) == "" {
-		return errors.New("empty copilot session token")
-	}
-	// Parse JWT exp
-	exp := parseJWTExp(out.Token)
-	if exp.IsZero() {
-		exp = time.Now().Add(10 * time.Minute)
-	}
-	c.sessionToken = out.Token
-	c.tokenExpiry = exp
-	return nil
-}
-
-var jwtExpRe = regexp.MustCompile(`"exp"\s*:\s*([0-9]+)`) // fallback if we can't base64 decode
-
-func parseJWTExp(token string) time.Time {
-	parts := strings.Split(token, ".")
-	if len(parts) < 2 {
-		return time.Time{}
-	}
-	b, err := base64.RawURLEncoding.DecodeString(parts[1])
-	if err != nil {
-		if m := jwtExpRe.FindStringSubmatch(token); len(m) == 2 {
-			if n, err2 := parseInt64(m[1]); err2 == nil {
-				return time.Unix(n, 0)
-			}
-		}
-		return time.Time{}
-	}
-	var payload struct {
-		Exp int64 `json:"exp"`
-	}
-	_ = json.Unmarshal(b, &payload)
-	if payload.Exp == 0 {
-		return time.Time{}
-	}
-	return time.Unix(payload.Exp, 0)
-}
-
-func parseInt64(s string) (int64, error) { var n int64; _, err := fmt.Sscan(s, &n); return n, err }
-
-// --- Copilot headers ---
-
-func (c *copilotClient) headersChat() map[string]string {
-	_ = c.ensureSession(context.Background())
-	h := map[string]string{
-		"Content-Type":          "application/json; charset=utf-8",
-		"Accept":                "application/json",
-		"Authorization":         "Bearer " + c.sessionToken,
-		"User-Agent":            "GitHubCopilotChat/0.8.0",
-		"Editor-Plugin-Version": "copilot-chat/0.8.0",
-		"Editor-Version":        "vscode/1.85.1",
-		"Openai-Intent":         "conversation-panel",
-		"Openai-Organization":   "github-copilot",
-		"VScode-MachineId":      randHex(64),
-		"VScode-SessionId":      randHex(8) + "-" + randHex(4) + "-" + randHex(4) + "-" + randHex(4) + "-" + randHex(12),
-		"X-Request-Id":          randHex(8) + "-" + randHex(4) + "-" + randHex(4) + "-" + randHex(4) + "-" + randHex(12),
-	}
-	return h
-}
-
-func (c *copilotClient) headersGhost() map[string]string {
-	_ = c.ensureSession(context.Background())
-	h := map[string]string{
-		"Content-Type":          "application/json; charset=utf-8",
-		"Accept":                "*/*",
-		"Authorization":         "Bearer " + c.sessionToken,
-		"User-Agent":            "GithubCopilot/1.155.0",
-		"Editor-Plugin-Version": "copilot/1.155.0",
-		"Editor-Version":        "vscode/1.85.1",
-		"Openai-Intent":         "copilot-ghost",
-		"Openai-Organization":   "github-copilot",
-		"VScode-MachineId":      randHex(64),
-		"VScode-SessionId":      randHex(8) + "-" + randHex(4) + "-" + randHex(4) + "-" + randHex(4) + "-" + randHex(12),
-		"X-Request-Id":          randHex(8) + "-" + randHex(4) + "-" + randHex(4) + "-" + randHex(4) + "-" + randHex(12),
-	}
-	return h
-}
-
-func randHex(n int) string {
-	const hex = "0123456789abcdef"
-	b := make([]byte, n)
-	for i := range b {
-		b[i] = hex[int(time.Now().UnixNano()+int64(i))%len(hex)]
-	}
-	return string(b)
-}
-
-// --- Codex-style code completion ---
-
-// CodeCompletion implements CodeCompleter; returns up to n suggestions.
-func (c copilotClient) CodeCompletion(ctx context.Context, prompt string, suffix string, n int, language string, temperature float64) ([]string, error) {
-	if strings.TrimSpace(c.apiKey) == "" {
-		return nil, errors.New("missing Copilot API key")
-	}
-	if err := c.ensureSession(ctx); err != nil {
-		return nil, err
-	}
-	if n <= 0 {
-		n = 1
-	}
-	maxTokens := 500
-	body := map[string]any{
-		"extra": map[string]any{
-			"language":            language,
-			"next_indent":         0,
-			"prompt_tokens":       500,
-			"suffix_tokens":       400,
-			"trim_by_indentation": true,
-		},
-		"max_tokens":  maxTokens,
-		"n":           n,
-		"nwo":         "hexai",
-		"prompt":      prompt,
-		"stop":        []string{"\n\n"},
-		"stream":      true,
-		"suffix":      suffix,
-		"temperature": temperature,
-		"top_p":       1,
-	}
-	buf, _ := json.Marshal(body)
-	url := "https://copilot-proxy.githubusercontent.com/v1/engines/copilot-codex/completions"
-	resp, err := c.postJSON(ctx, url, buf, c.headersGhost())
-	if err != nil {
-		return nil, err
-	}
-	defer func() {
-		if err := resp.Body.Close(); err != nil {
-			logging.Logf("llm/copilot", "failed to close response body: %v", err)
-		}
-	}()
-	if resp.StatusCode < 200 || resp.StatusCode >= 300 {
-		return nil, fmt.Errorf("copilot codex http error: %d", resp.StatusCode)
-	}
-	// Read all and parse lines that start with "data: " accumulating by index
-	raw, _ := io.ReadAll(resp.Body)
-	byIndex := make(map[int]string)
-	lines := strings.Split(string(raw), "\n")
-	for _, ln := range lines {
-		if !strings.HasPrefix(ln, "data: ") {
-			continue
-		}
-		var evt struct {
-			Choices []struct {
-				Index int    `json:"index"`
-				Text  string `json:"text"`
-			} `json:"choices"`
-		}
-		if err := json.Unmarshal([]byte(strings.TrimPrefix(ln, "data: ")), &evt); err != nil {
-			continue
-		}
-		for _, ch := range evt.Choices {
-			byIndex[ch.Index] += ch.Text
-		}
-	}
-	out := make([]string, 0, len(byIndex))
-	for i := 0; i < n; i++ {
-		if s, ok := byIndex[i]; ok && strings.TrimSpace(s) != "" {
-			out = append(out, s)
-		}
-	}
-	return out, nil
-}
-
-// newLineDataReader wraps a streaming body and exposes a JSON decoder that
-// decodes successive objects from lines prefixed by "data: ".
-// (no streaming decoder needed; we parse whole body lines)
diff --git a/internal/llm/copilot_http_test.go b/internal/llm/copilot_http_test.go
deleted file mode 100644
index 1371f71..0000000
--- a/internal/llm/copilot_http_test.go
+++ /dev/null
@@ -1,276 +0,0 @@
-package llm
-
-import (
-	"context"
-	"encoding/base64"
-	"encoding/json"
-	"io"
-	"net"
-	"net/http"
-	"net/http/httptest"
-	"os"
-	"strings"
-	"testing"
-	"time"
-)
-
-type rtFunc2 func(*http.Request) (*http.Response, error)
-
-func (f rtFunc2) RoundTrip(r *http.Request) (*http.Response, error) { return f(r) }
-
-func TestCopilot_EnsureSession_AndChat_Success(t *testing.T) {
-	if os.Getenv("HEXAI_TEST_SKIP_NET") == "1" {
-		t.Skip("skip network-bound tests in restricted environments")
-	}
-	// Mock chat endpoint
-	chatSrv := newIPv4Server(t, http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		if r.URL.Path != "/chat/completions" {
-			t.Fatalf("unexpected path: %s", r.URL.Path)
-		}
-		_ = json.NewEncoder(w).Encode(map[string]any{"choices": []map[string]any{{"index": 0, "message": map[string]string{"role": "assistant", "content": "OK"}}}})
-	}))
-	defer chatSrv.Close()
-	c := newCopilot(chatSrv.URL, "gpt-4o-mini", "APIKEY", f64p(0.1)).(copilotClient)
-	// Intercept token endpoint to return a session token
-	tr := rtFunc2(func(r *http.Request) (*http.Response, error) {
-		if r.URL.Host == "api.github.com" && r.URL.Path == "/copilot_internal/v2/token" {
-			rw := httptest.NewRecorder()
-			_ = json.NewEncoder(rw).Encode(map[string]string{"token": "tok"})
-			res := rw.Result()
-			res.StatusCode = 200
-			return res, nil
-		}
-		// Fallback to default transport for chatSrv
-		return http.DefaultTransport.RoundTrip(r)
-	})
-	c.httpClient = &http.Client{Transport: tr, Timeout: 5 * time.Second}
-	out, err := c.Chat(context.Background(), []Message{{Role: "user", Content: "hi"}})
-	if err != nil || out != "OK" {
-		t.Fatalf("copilot chat failed: %v %q", err, out)
-	}
-}
-
-func TestCopilot_HandleNon2xx(t *testing.T) {
-	b, _ := json.Marshal(map[string]any{"error": map[string]any{"message": "bad", "type": "invalid"}})
-	resp := &http.Response{StatusCode: 400, Body: io.NopCloser(bytesReader(b))}
-	if err := handleCopilotNon2xx(resp, time.Now()); err == nil {
-		t.Fatalf("expected error")
-	}
-}
-
-func TestCopilot_CodeCompletion_Success(t *testing.T) {
-	c := newCopilot("https://api.githubcopilot.com", "gpt-4o-mini", "API", f64p(0.1)).(copilotClient)
-	tr := rtFunc2(func(r *http.Request) (*http.Response, error) {
-		// Token endpoint
-		if r.URL.Host == "api.github.com" && r.URL.Path == "/copilot_internal/v2/token" {
-			rw := httptest.NewRecorder()
-			_ = json.NewEncoder(rw).Encode(map[string]string{"token": "tok"})
-			res := rw.Result()
-			res.StatusCode = 200
-			return res, nil
-		}
-		// Codex completion endpoint
-		if r.URL.Host == "copilot-proxy.githubusercontent.com" && strings.HasSuffix(r.URL.Path, "/v1/engines/copilot-codex/completions") {
-			rw := httptest.NewRecorder()
-			// two choices for index 0 and 1
-			_, _ = rw.WriteString("data: {\"choices\":[{\"index\":0,\"text\":\"A\"}]}\n")
-			_, _ = rw.WriteString("data: {\"choices\":[{\"index\":1,\"text\":\"B\"}]}\n")
-			res := rw.Result()
-			res.StatusCode = 200
-			return res, nil
-		}
-		return http.DefaultTransport.RoundTrip(r)
-	})
-	c.httpClient = &http.Client{Transport: tr, Timeout: 5 * time.Second}
-	out, err := c.CodeCompletion(context.Background(), "p", "s", 2, "go", 0.1)
-	if err != nil || len(out) != 2 || out[0] != "A" || out[1] != "B" {
-		t.Fatalf("codex: %v %#v", err, out)
-	}
-}
-
-func TestCopilot_Chat_MultiChoice_And_ErrorBody(t *testing.T) {
-	if os.Getenv("HEXAI_TEST_SKIP_NET") == "1" {
-		t.Skip("skip network-bound tests in restricted environments")
-	}
-	// Chat multi-choice: return two choices; client returns first content
-	srv := newIPv4Server(t, http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		_ = json.NewEncoder(w).Encode(map[string]any{
-			"choices": []map[string]any{
-				{"index": 0, "finish_reason": "stop", "message": map[string]string{"role": "assistant", "content": "FIRST"}},
-				{"index": 1, "finish_reason": "length", "message": map[string]string{"role": "assistant", "content": "SECOND"}},
-			},
-		})
-	}))
-	defer srv.Close()
-	c := newCopilot(srv.URL, "gpt-4o-mini", "KEY", f64p(0.1)).(copilotClient)
-	// Token success
-	tr := rtFunc2(func(r *http.Request) (*http.Response, error) {
-		if r.URL.Host == "api.github.com" && r.URL.Path == "/copilot_internal/v2/token" {
-			rw := httptest.NewRecorder()
-			_ = json.NewEncoder(rw).Encode(map[string]string{"token": "tok"})
-			res := rw.Result()
-			res.StatusCode = 200
-			return res, nil
-		}
-		return http.DefaultTransport.RoundTrip(r)
-	})
-	c.httpClient = &http.Client{Transport: tr, Timeout: 5 * time.Second}
-	out, err := c.Chat(context.Background(), []Message{{Role: "user", Content: "hi"}})
-	if err != nil || out != "FIRST" {
-		t.Fatalf("copilot multi-choice: %v %q", err, out)
-	}
-
-	// Non-2xx with error body
-	srv2 := newIPv4Server(t, http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		w.WriteHeader(403)
-		_ = json.NewEncoder(w).Encode(map[string]any{"error": map[string]any{"message": "denied", "type": "forbidden"}})
-	}))
-	defer srv2.Close()
-	c2 := newCopilot(srv2.URL, "gpt-4o-mini", "KEY", f64p(0.1)).(copilotClient)
-	c2.httpClient = &http.Client{Transport: tr, Timeout: 5 * time.Second}
-	if _, err := c2.Chat(context.Background(), []Message{{Role: "user", Content: "hi"}}); err == nil {
-		t.Fatalf("expected error for copilot non-2xx with error body")
-	}
-}
-
-func TestCopilot_Chat_NoChoices_Error(t *testing.T) {
-	if os.Getenv("HEXAI_TEST_SKIP_NET") == "1" {
-		t.Skip("skip network-bound tests in restricted environments")
-	}
-	srv := newIPv4Server(t, http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		_ = json.NewEncoder(w).Encode(map[string]any{"choices": []any{}})
-	}))
-	defer srv.Close()
-	c := newCopilot(srv.URL, "gpt-4o-mini", "KEY", f64p(0.1)).(copilotClient)
-	tr := rtFunc2(func(r *http.Request) (*http.Response, error) {
-		if r.URL.Host == "api.github.com" && r.URL.Path == "/copilot_internal/v2/token" {
-			rw := httptest.NewRecorder()
-			_ = json.NewEncoder(rw).Encode(map[string]string{"token": "tok"})
-			res := rw.Result()
-			res.StatusCode = 200
-			return res, nil
-		}
-		return http.DefaultTransport.RoundTrip(r)
-	})
-	c.httpClient = &http.Client{Transport: tr, Timeout: 5 * time.Second}
-	if _, err := c.Chat(context.Background(), []Message{{Role: "user", Content: "hi"}}); err == nil {
-		t.Fatalf("expected error when no choices returned")
-	}
-}
-
-func TestCopilot_Chat_DecodeError_StatusOK(t *testing.T) {
-	if os.Getenv("HEXAI_TEST_SKIP_NET") == "1" {
-		t.Skip("skip network-bound tests in restricted environments")
-	}
-	// Chat returns 200 but invalid JSON; expect decode error
-	srv := newIPv4Server(t, http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		_, _ = io.WriteString(w, "{invalid")
-	}))
-	defer srv.Close()
-	c := newCopilot(srv.URL, "gpt-4o-mini", "KEY", f64p(0.1)).(copilotClient)
-	tr := rtFunc2(func(r *http.Request) (*http.Response, error) {
-		if r.URL.Host == "api.github.com" && r.URL.Path == "/copilot_internal/v2/token" {
-			rw := httptest.NewRecorder()
-			_ = json.NewEncoder(rw).Encode(map[string]string{"token": "tok"})
-			res := rw.Result()
-			res.StatusCode = 200
-			return res, nil
-		}
-		return http.DefaultTransport.RoundTrip(r)
-	})
-	c.httpClient = &http.Client{Transport: tr, Timeout: 5 * time.Second}
-	if _, err := c.Chat(context.Background(), []Message{{Role: "user", Content: "hi"}}); err == nil {
-		t.Fatalf("expected decode error for invalid body")
-	}
-}
-
-func TestCopilot_CodeCompletion_MalformedAndEmpty(t *testing.T) {
-	c := newCopilot("https://api.githubcopilot.com", "gpt-4o-mini", "API", f64p(0.1)).(copilotClient)
-	tr := rtFunc2(func(r *http.Request) (*http.Response, error) {
-		if r.URL.Host == "api.github.com" && r.URL.Path == "/copilot_internal/v2/token" {
-			rw := httptest.NewRecorder()
-			_ = json.NewEncoder(rw).Encode(map[string]string{"token": "tok"})
-			res := rw.Result()
-			res.StatusCode = 200
-			return res, nil
-		}
-		if r.URL.Host == "copilot-proxy.githubusercontent.com" && strings.HasSuffix(r.URL.Path, "/v1/engines/copilot-codex/completions") {
-			rw := httptest.NewRecorder()
-			// malformed line
-			_, _ = rw.WriteString("data: {bad}\n")
-			// done; should produce empty suggestions
-			_, _ = rw.WriteString("data: [DONE]\n")
-			res := rw.Result()
-			res.StatusCode = 200
-			return res, nil
-		}
-		return http.DefaultTransport.RoundTrip(r)
-	})
-	c.httpClient = &http.Client{Transport: tr, Timeout: 5 * time.Second}
-	out, err := c.CodeCompletion(context.Background(), "p", "s", 1, "go", 0.1)
-	if err != nil {
-		t.Fatalf("unexpected error: %v", err)
-	}
-	if len(out) != 0 {
-		t.Fatalf("expected empty suggestions, got %#v", out)
-	}
-
-	// Now include one good chunk after malformed
-	tr2 := rtFunc2(func(r *http.Request) (*http.Response, error) {
-		if r.URL.Host == "api.github.com" && r.URL.Path == "/copilot_internal/v2/token" {
-			rw := httptest.NewRecorder()
-			_ = json.NewEncoder(rw).Encode(map[string]string{"token": "tok"})
-			res := rw.Result()
-			res.StatusCode = 200
-			return res, nil
-		}
-		if r.URL.Host == "copilot-proxy.githubusercontent.com" && strings.HasSuffix(r.URL.Path, "/v1/engines/copilot-codex/completions") {
-			rw := httptest.NewRecorder()
-			_, _ = rw.WriteString("data: {bad}\n")
-			_, _ = rw.WriteString("data: {\"choices\":[{\"index\":0,\"text\":\"OK\"}]}\n")
-			_, _ = rw.WriteString("data: [DONE]\n")
-			res := rw.Result()
-			res.StatusCode = 200
-			return res, nil
-		}
-		return http.DefaultTransport.RoundTrip(r)
-	})
-	c.httpClient = &http.Client{Transport: tr2, Timeout: 5 * time.Second}
-	out2, err := c.CodeCompletion(context.Background(), "p", "s", 1, "go", 0.1)
-	if err != nil || len(out2) != 1 || out2[0] != "OK" {
-		t.Fatalf("unexpected: %v %#v", err, out2)
-	}
-}
-
-func TestParseJWTExp_AndParseInt64(t *testing.T) {
-	// Valid base64 payload
-	payload := `{"exp": 1700000000}`
-	b := base64.RawURLEncoding.EncodeToString([]byte(payload))
-	tok := "x." + b + ".y"
-	if tm := parseJWTExp(tok); tm.IsZero() {
-		t.Fatalf("expected non-zero time")
-	}
-	if n, err := parseInt64("123"); err != nil || n != 123 {
-		t.Fatalf("parseInt64: %v %d", err, n)
-	}
-}
-
-func newIPv4Server(t *testing.T, handler http.Handler) *httptest.Server {
-	t.Helper()
-	l, err := net.Listen("tcp4", "127.0.0.1:0")
-	if err != nil {
-		t.Fatalf("failed to listen on tcp4: %v", err)
-	}
-	srv := &httptest.Server{
-		Listener: l,
-		Config:   &http.Server{Handler: handler},
-	}
-	srv.Start()
-	return srv
-}
-
-// bytesReader wraps a byte slice with an io.ReadCloser without importing extra.
-type bytesReader []byte
-
-func (b bytesReader) Read(p []byte) (int, error) { n := copy(p, b); return n, io.EOF }
-func (b bytesReader) Close() error               { return nil }
diff --git a/internal/llm/copilot_test.go b/internal/llm/copilot_test.go
deleted file mode 100644
index 8f15347..0000000
--- a/internal/llm/copilot_test.go
+++ /dev/null
@@ -1,35 +0,0 @@
-package llm
-
-import "testing"
-
-func TestBuildCopilotChatRequest_FieldsAndDefaults(t *testing.T) {
-	o := Options{
-		Model:       "gpt-x",
-		Temperature: 0,
-		MaxTokens:   123,
-		Stop:        []string{"X"},
-	}
-
-	msgs := []Message{{Role: "user", Content: "q"}}
-	req := buildCopilotChatRequest(o, msgs, f64p(0.5))
-
-	if req.Model != "gpt-x" {
-		t.Fatalf("model mismatch: %q", req.Model)
-	}
-
-	if req.Temperature == nil || *req.Temperature != 0.5 {
-		t.Fatalf("default temp not applied")
-	}
-
-	if req.MaxTokens == nil || *req.MaxTokens != 123 {
-		t.Fatalf("max_tokens not applied")
-	}
-
-	if len(req.Stop) != 1 || req.Stop[0] != "X" {
-		t.Fatalf("stop not applied")
-	}
-
-	if len(req.Messages) != 1 || req.Messages[0].Content != "q" {
-		t.Fatalf("messages not copied")
-	}
-}
diff --git a/internal/llm/openai_temp_test.go b/internal/llm/openai_temp_test.go
index 07abbd5..3d71b94 100644
--- a/internal/llm/openai_temp_test.go
+++ b/internal/llm/openai_temp_test.go
@@ -5,7 +5,7 @@ import "testing"
 func TestNewFromConfig_DefaultTemp_ByModel(t *testing.T) {
 	// OpenAI, gpt-5.* → default temp 1.0 when not provided
 	cfg := Config{Provider: "openai", OpenAIModel: "gpt-5.0-preview"}
-	c, err := NewFromConfig(cfg, "key", "", "", "")
+	c, err := NewFromConfig(cfg, "key", "", "")
 	if err != nil {
 		t.Fatalf("new: %v", err)
 	}
@@ -18,7 +18,7 @@ func TestNewFromConfig_DefaultTemp_ByModel(t *testing.T) {
 	}
 	// OpenAI, gpt-4.* → default temp 0.2 when not provided
 	cfg2 := Config{Provider: "openai", OpenAIModel: "gpt-4.1"}
-	c2, err := NewFromConfig(cfg2, "key", "", "", "")
+	c2, err := NewFromConfig(cfg2, "key", "", "")
 	if err != nil {
 		t.Fatalf("new2: %v", err)
 	}
@@ -32,7 +32,7 @@ func TestNewFromConfig_DefaultTemp_UpgradeWhenGpt5AndDefault02(t *testing.T) {
 	// Simulate app-default of 0.2 while selecting a gpt-5 model: should upgrade to 1.0
 	v := 0.2
 	cfg := Config{Provider: "openai", OpenAIModel: "gpt-5.0", OpenAITemperature: &v}
-	c, err := NewFromConfig(cfg, "key", "", "", "")
+	c, err := NewFromConfig(cfg, "key", "", "")
 	if err != nil {
 		t.Fatalf("new: %v", err)
 	}
diff --git a/internal/llm/provider.go b/internal/llm/provider.go
index 297f1f3..8230b53 100644
--- a/internal/llm/provider.go
+++ b/internal/llm/provider.go
@@ -78,10 +78,6 @@ type Config struct {
 	OllamaBaseURL     string
 	OllamaModel       string
 	OllamaTemperature *float64
-	// Copilot options
-	CopilotBaseURL     string
-	CopilotModel       string
-	CopilotTemperature *float64
 	// Anthropic options
 	AnthropicBaseURL     string
 	AnthropicModel       string
@@ -91,7 +87,7 @@ type Config struct {
 // NewFromConfig creates an LLM client using only the supplied configuration.
 // The OpenAI API key is supplied separately and may be read from the environment
 // by the caller; other environment-based configuration is not used.
-func NewFromConfig(cfg Config, openAIAPIKey, openRouterAPIKey, copilotAPIKey, anthropicAPIKey string) (Client, error) {
+func NewFromConfig(cfg Config, openAIAPIKey, openRouterAPIKey, anthropicAPIKey string) (Client, error) {
 	p := strings.ToLower(strings.TrimSpace(cfg.Provider))
 	if p == "" {
 		p = "openai"
@@ -136,15 +132,6 @@ func NewFromConfig(cfg Config, openAIAPIKey, openRouterAPIKey, copilotAPIKey, an
 			cfg.OllamaTemperature = &t
 		}
 		return newOllamaWithTimeout(cfg.OllamaBaseURL, cfg.OllamaModel, cfg.OllamaTemperature, cfg.RequestTimeout), nil
-	case "copilot":
-		if strings.TrimSpace(copilotAPIKey) == "" {
-			return nil, errors.New("missing COPILOT_API_KEY for provider copilot")
-		}
-		if cfg.CopilotTemperature == nil {
-			t := 0.2
-			cfg.CopilotTemperature = &t
-		}
-		return newCopilotWithTimeout(cfg.CopilotBaseURL, cfg.CopilotModel, copilotAPIKey, cfg.CopilotTemperature, cfg.RequestTimeout), nil
 	case "anthropic":
 		if strings.TrimSpace(anthropicAPIKey) == "" {
 			return nil, errors.New("missing ANTHROPIC_API_KEY for provider anthropic")
diff --git a/internal/llm/provider_more2_test.go b/internal/llm/provider_more2_test.go
deleted file mode 100644
index 86b149a..0000000
--- a/internal/llm/provider_more2_test.go
+++ /dev/null
@@ -1,12 +0,0 @@
-package llm
-
-import "testing"
-
-func TestNewFromConfig_Copilot(t *testing.T) {
-	t.Setenv("COPILOT_API_KEY", "x")
-	cfg := Config{Provider: "copilot", CopilotModel: "small"}
-	c, err := NewFromConfig(cfg, "", "", "x", "")
-	if err != nil || c == nil {
-		t.Fatalf("copilot provider failed: %v %v", c, err)
-	}
-}
diff --git a/internal/llm/provider_more_test.go b/internal/llm/provider_more_test.go
index caad912..8d7b133 100644
--- a/internal/llm/provider_more_test.go
+++ b/internal/llm/provider_more_test.go
@@ -13,17 +13,11 @@ func TestWithOptions_Apply(t *testing.T) {
 	}
 }
 
-func TestNewFromConfig_Success_OpenAI_And_Copilot(t *testing.T) {
+func TestNewFromConfig_Success_OpenAI(t *testing.T) {
 	// OpenAI success
 	oc := Config{Provider: "openai", OpenAIBaseURL: "http://x", OpenAIModel: "gpt"}
-	c, err := NewFromConfig(oc, "KEY", "", "", "")
+	c, err := NewFromConfig(oc, "KEY", "", "")
 	if err != nil || c == nil || c.Name() != "openai" || c.DefaultModel() == "" {
 		t.Fatalf("openai new: %v %v", c, err)
 	}
-	// Copilot success
-	cc := Config{Provider: "copilot", CopilotBaseURL: "http://x", CopilotModel: "gpt-4o-mini"}
-	c2, err := NewFromConfig(cc, "", "", "KEY", "")
-	if err != nil || c2 == nil || c2.Name() != "copilot" || c2.DefaultModel() == "" {
-		t.Fatalf("copilot new: %v %v", c2, err)
-	}
 }
diff --git a/internal/llm/provider_test.go b/internal/llm/provider_test.go
index 46c7ea8..8ccba6e 100644
--- a/internal/llm/provider_test.go
+++ b/internal/llm/provider_test.go
@@ -6,15 +6,11 @@ import (
 
 func TestNewFromConfig_DefaultsAndErrors(t *testing.T) {
 	// Unknown provider
-	if _, err := NewFromConfig(Config{Provider: "bogus"}, "", "", "", ""); err == nil {
+	if _, err := NewFromConfig(Config{Provider: "bogus"}, "", "", ""); err == nil {
 		t.Fatalf("expected error for unknown provider")
 	}
 	// OpenAI missing key
-	if _, err := NewFromConfig(Config{Provider: "openai", OpenAIModel: "g"}, "", "", "", ""); err == nil {
-		t.Fatalf("expected key error")
-	}
-	// Copilot missing key
-	if _, err := NewFromConfig(Config{Provider: "copilot", CopilotModel: "m"}, "", "", "", ""); err == nil {
+	if _, err := NewFromConfig(Config{Provider: "openai", OpenAIModel: "g"}, "", "", ""); err == nil {
 		t.Fatalf("expected key error")
 	}
 }
diff --git a/internal/llmutils/client.go b/internal/llmutils/client.go
index de65935..c8d9a90 100644
--- a/internal/llmutils/client.go
+++ b/internal/llmutils/client.go
@@ -22,9 +22,6 @@ func NewClientFromApp(cfg appconfig.App) (llm.Client, error) {
 		OllamaBaseURL:         cfg.OllamaBaseURL,
 		OllamaModel:           cfg.OllamaModel,
 		OllamaTemperature:     cfg.OllamaTemperature,
-		CopilotBaseURL:        cfg.CopilotBaseURL,
-		CopilotModel:          cfg.CopilotModel,
-		CopilotTemperature:    cfg.CopilotTemperature,
 		AnthropicBaseURL:      cfg.AnthropicBaseURL,
 		AnthropicModel:        cfg.AnthropicModel,
 		AnthropicTemperature:  cfg.AnthropicTemperature,
@@ -37,13 +34,9 @@ func NewClientFromApp(cfg appconfig.App) (llm.Client, error) {
 	if strings.TrimSpace(orKey) == "" {
 		orKey = os.Getenv("OPENROUTER_API_KEY")
 	}
-	cpKey := os.Getenv("HEXAI_COPILOT_API_KEY")
-	if strings.TrimSpace(cpKey) == "" {
-		cpKey = os.Getenv("COPILOT_API_KEY")
-	}
 	anKey := os.Getenv("HEXAI_ANTHROPIC_API_KEY")
 	if strings.TrimSpace(anKey) == "" {
 		anKey = os.Getenv("ANTHROPIC_API_KEY")
 	}
-	return llm.NewFromConfig(llmCfg, oaKey, orKey, cpKey, anKey)
+	return llm.NewFromConfig(llmCfg, oaKey, orKey, anKey)
 }
diff --git a/internal/lsp/handlers_utils.go b/internal/lsp/handlers_utils.go
index 6260acd..1ea36c8 100644
--- a/internal/lsp/handlers_utils.go
+++ b/internal/lsp/handlers_utils.go
@@ -110,8 +110,6 @@ func resolveDefaultModel(cfg appconfig.App, provider string) string {
 	switch provider {
 	case "ollama":
 		return strings.TrimSpace(cfg.OllamaModel)
-	case "copilot":
-		return strings.TrimSpace(cfg.CopilotModel)
 	case "anthropic":
 		return strings.TrimSpace(cfg.AnthropicModel)
 	case "openrouter":
diff --git a/internal/lsp/llm_request_opts_test.go b/internal/lsp/llm_request_opts_test.go
index f4d31b9..ad87cd4 100644
--- a/internal/lsp/llm_request_opts_test.go
+++ b/internal/lsp/llm_request_opts_test.go
@@ -44,10 +44,10 @@ func TestBuildRequestSpecs_MultiEntries(t *testing.T) {
 	s := newTestServer()
 	s.cfg.CompletionConfigs = []appconfig.SurfaceConfig{
 		{Provider: "openai", Model: "gpt-4o"},
-		{Provider: "copilot", Model: "cpt", Temperature: floatPtr(0.4)},
+		{Provider: "anthropic", Model: "claude", Temperature: floatPtr(0.4)},
 	}
 	s.cfg.OpenAIModel = "gpt-3.5"
-	s.cfg.CopilotModel = "cpt-base"
+	s.cfg.AnthropicModel = "claude-base"
 	s.cfg.MaxTokens = 256
 	specs := s.buildRequestSpecs(surfaceCompletion)
 	if len(specs) != 2 {
@@ -56,7 +56,7 @@ func TestBuildRequestSpecs_MultiEntries(t *testing.T) {
 	if specs[0].provider != "openai" || specs[0].index != 0 {
 		t.Fatalf("unexpected first spec: %+v", specs[0])
 	}
-	if specs[1].provider != "copilot" || specs[1].index != 1 {
+	if specs[1].provider != "anthropic" || specs[1].index != 1 {
 		t.Fatalf("unexpected second spec: %+v", specs[1])
 	}
 	var opts1, opts2 llm.Options
@@ -69,7 +69,7 @@ func TestBuildRequestSpecs_MultiEntries(t *testing.T) {
 	if opts1.Model != "gpt-4o" || opts1.MaxTokens != 256 {
 		t.Fatalf("unexpected opts1: %+v", opts1)
 	}
-	if opts2.Model != "cpt" || opts2.Temperature != 0.4 {
+	if opts2.Model != "claude" || opts2.Temperature != 0.4 {
 		t.Fatalf("unexpected opts2: %+v", opts2)
 	}
 }
diff --git a/internal/lsp/server.go b/internal/lsp/server.go
index bbee64f..c226ab4 100644
--- a/internal/lsp/server.go
+++ b/internal/lsp/server.go
@@ -230,9 +230,6 @@ func newClientForProvider(cfg appconfig.App, provider string) (llm.Client, error
 		OllamaBaseURL:         cfg.OllamaBaseURL,
 		OllamaModel:           cfg.OllamaModel,
 		OllamaTemperature:     cfg.OllamaTemperature,
-		CopilotBaseURL:        cfg.CopilotBaseURL,
-		CopilotModel:          cfg.CopilotModel,
-		CopilotTemperature:    cfg.CopilotTemperature,
 		AnthropicBaseURL:      cfg.AnthropicBaseURL,
 		AnthropicModel:        cfg.AnthropicModel,
 		AnthropicTemperature:  cfg.AnthropicTemperature,
@@ -245,15 +242,11 @@ func newClientForProvider(cfg appconfig.App, provider string) (llm.Client, error
 	if orKey == "" {
 		orKey = strings.TrimSpace(os.Getenv("OPENROUTER_API_KEY"))
 	}
-	cpKey := strings.TrimSpace(os.Getenv("HEXAI_COPILOT_API_KEY"))
-	if cpKey == "" {
-		cpKey = strings.TrimSpace(os.Getenv("COPILOT_API_KEY"))
-	}
 	anKey := strings.TrimSpace(os.Getenv("HEXAI_ANTHROPIC_API_KEY"))
 	if anKey == "" {
 		anKey = strings.TrimSpace(os.Getenv("ANTHROPIC_API_KEY"))
 	}
-	return llm.NewFromConfig(llmCfg, oaKey, orKey, cpKey, anKey)
+	return llm.NewFromConfig(llmCfg, oaKey, orKey, anKey)
 }
 
 func (s *Server) clientFor(spec requestSpec) llm.Client {
@@ -296,12 +289,6 @@ func (s *Server) clientFor(spec requestSpec) llm.Client {
 		} else if spec.fallbackModel != "" {
 			cfg.OpenRouterModel = spec.fallbackModel
 		}
-	case "copilot":
-		if modelOverride != "" {
-			cfg.CopilotModel = modelOverride
-		} else if spec.fallbackModel != "" {
-			cfg.CopilotModel = spec.fallbackModel
-		}
 	case "ollama":
 		if modelOverride != "" {
 			cfg.OllamaModel = modelOverride
diff --git a/internal/runtimeconfig/store_test.go b/internal/runtimeconfig/store_test.go
index 0a0183a..168d2cd 100644
--- a/internal/runtimeconfig/store_test.go
+++ b/internal/runtimeconfig/store_test.go
@@ -99,7 +99,7 @@ func TestStoreReloadLogsSummary(t *testing.T) {
 
 func TestDiff_SurfaceModel(t *testing.T) {
 	oldCfg := appconfig.App{CompletionConfigs: []appconfig.SurfaceConfig{{Provider: "openai", Model: "gpt-4o"}}}
-	newCfg := appconfig.App{CompletionConfigs: []appconfig.SurfaceConfig{{Provider: "copilot", Model: "gpt-4.1"}}}
+	newCfg := appconfig.App{CompletionConfigs: []appconfig.SurfaceConfig{{Provider: "anthropic", Model: "claude-3-5-sonnet"}}}
 	changes := Diff(oldCfg, newCfg)
 	if len(changes) == 0 {
 		t.Fatalf("expected diff entries, got none")
@@ -107,7 +107,7 @@ func TestDiff_SurfaceModel(t *testing.T) {
 	found := false
 	for _, ch := range changes {
 		if ch.Key == "completion_configs" {
-			if !strings.Contains(ch.Old, "gpt-4o") || !strings.Contains(ch.New, "gpt-4.1") {
+			if !strings.Contains(ch.Old, "gpt-4o") || !strings.Contains(ch.New, "claude-3-5-sonnet") {
 				t.Fatalf("unexpected diff contents: %+v", ch)
 			}
 			found = true