diff --git a/apps/showcase_ui/src/app/pages/home/home.component.spec.ts b/apps/showcase_ui/src/app/pages/home/home.component.spec.ts
index 71bc830..30da154 100644
--- a/apps/showcase_ui/src/app/pages/home/home.component.spec.ts
+++ b/apps/showcase_ui/src/app/pages/home/home.component.spec.ts
@@ -10,7 +10,7 @@ import { TaskRecommendationService } from '../../core/services/task-recommendati
describe('Homepage exploration entry',()=>{
let api:any,agent:any,router:any,system:any;
- const config={model:'gpt-6.1-sol',reasoning_effort:'high',timeout_seconds:1800,demonstration_timeout_seconds:900,demonstration_warning_seconds:60};
+ const config={model:'gpt-6.1-terra',reasoning_effort:'high',timeout_seconds:1800,demonstration_timeout_seconds:900,demonstration_warning_seconds:60};
beforeEach(()=>{
api={settings:jasmine.createSpy().and.resolveTo(config),create:jasmine.createSpy().and.resolveTo({id:'conversation'}),act:jasmine.createSpy().and.resolveTo({run:{id:'run'}})};
agent={runTask:jasmine.createSpy(),selectSession:jasmine.createSpy(),fetchSessions:jasmine.createSpy(),fetchStatus:jasmine.createSpy()};router={navigate:jasmine.createSpy().and.resolveTo(true)};
diff --git a/artemis/config/exploration.py b/artemis/config/exploration.py
index 451877b..76f96c3 100644
--- a/artemis/config/exploration.py
+++ b/artemis/config/exploration.py
@@ -8,7 +8,7 @@ from pydantic import BaseModel, ConfigDict, Field, model_validator
class ExplorationConfig(BaseModel):
model_config = ConfigDict(extra="forbid")
- model: str = Field(default="gpt-6.1-sol", min_length=1, max_length=200)
+ model: str = Field(default="gpt-6.1-terra", min_length=1, max_length=200)
reasoning_effort: Literal[
"default", "none", "minimal", "low", "medium", "high", "xhigh", "max"
] = "high"
diff --git a/artemis/resources/config/artemis.jsonc b/artemis/resources/config/artemis.jsonc
index 1387e2c..cd43142 100644
--- a/artemis/resources/config/artemis.jsonc
+++ b/artemis/resources/config/artemis.jsonc
@@ -35,7 +35,7 @@
},
"agent": {
"exploration": {
- "model": "${env:ARTEMIS_EXPLORATION_MODEL:-gpt-6.1-sol}",
+ "model": "${env:ARTEMIS_EXPLORATION_MODEL:-gpt-6.1-terra}",
"reasoning_effort": "${env:ARTEMIS_EXPLORATION_REASONING_EFFORT:-high}",
"timeout_seconds": "${env:ARTEMIS_EXPLORATION_TIMEOUT_SECONDS:-1800}",
"demonstration_timeout_seconds": "${env:ARTEMIS_DEMONSTRATION_TIMEOUT_SECONDS:-900}",
diff --git a/config/artemis.jsonc b/config/artemis.jsonc
index 1387e2c..cd43142 100644
--- a/config/artemis.jsonc
+++ b/config/artemis.jsonc
@@ -35,7 +35,7 @@
},
"agent": {
"exploration": {
- "model": "${env:ARTEMIS_EXPLORATION_MODEL:-gpt-6.1-sol}",
+ "model": "${env:ARTEMIS_EXPLORATION_MODEL:-gpt-6.1-terra}",
"reasoning_effort": "${env:ARTEMIS_EXPLORATION_REASONING_EFFORT:-high}",
"timeout_seconds": "${env:ARTEMIS_EXPLORATION_TIMEOUT_SECONDS:-1800}",
"demonstration_timeout_seconds": "${env:ARTEMIS_DEMONSTRATION_TIMEOUT_SECONDS:-900}",
diff --git a/docs/ai-automation.md b/docs/ai-automation.md
index 7a9c7be..4a2a5f5 100644
--- a/docs/ai-automation.md
+++ b/docs/ai-automation.md
@@ -60,7 +60,7 @@ MCP 同步仅检查 `$CODEX_HOME/config.toml`；未设置或为空时使用当
报告按任务或批次归档到 `artifacts/mcp_exports/`，包含本地路径及服务下载链接；删除批次时会清理对应导出副本。运行中导出是快照；执行完成不等于验证通过，报告会保留未验证、失败、跳过与中断状态。
-新安装模型默认使用 `gpt-6-luna`，预设 `gpt-6-luna`、`gpt-6.1-sol`、`gpt-6-astra`、`mimo-v2.6-pro`；首页、兜底模型与精选任务使用统一选择控件，也支持自定义模型名称。已保存的模型配置保留。思考强度最高支持 `max`，默认 `high`，选“由模型默认”不发送该参数；接口不支持时明确报错。单次请求 300 秒，首次失败后最多再试 5 次，普通错误每次等待 10 秒，429 每次等待 60 秒，SDK 不叠加重试。首选模型连续 3 次没有有效工具调用时，当前步骤改用 `gpt-6-luna`（high），最多再试 3 次；下一步重新使用首选模型。兜底模型及次数可通过 `.env.example` 中的 `OPENAI_TOOL_FALLBACK_*` 和 `ARTEMIS_TOOL_FALLBACK_MAX_NO_PROGRESS_ATTEMPTS` 配置。请求耗尽后沿用任务暂停/恢复机制，鉴权或非法参数立即失败。视频分析与录屏是独立开关，默认均关闭。每项批量用例可覆盖模式、模型、设备、录屏和超时等配置，不影响其他任务。暂停先返回 `pausing`，确认状态为 `task_paused` 才表示设备已释放；Pro 等待模型回复或只读诊断时可及时中断，恢复后重新观察设备。AI 可用 `mobile_manage_task(action="inject_instruction")` 提交指导并恢复；指导由执行器在安全检查点领取，可能中断当前模型等待并重新观察设备；接收回执不保证任务结束前已消费，已经开始的设备动作会安全结束。用 `guidance_history` 核对指导是否被领取、在任务流程中核对实际执行情况，或用 `resume` 单独恢复。
+新安装模型默认使用 `gpt-6-luna`，预设 `gpt-6-luna`、`gpt-6.1-terra`、`gpt-6.1-sol`、`gpt-6-astra`、`mimo-v2.6-pro`；首页、兜底模型与精选任务使用统一选择控件，也支持自定义模型名称。已保存的模型配置保留。思考强度最高支持 `max`，默认 `high`，选“由模型默认”不发送该参数；接口不支持时明确报错。单次请求 300 秒，首次失败后最多再试 5 次，普通错误每次等待 10 秒，429 每次等待 60 秒，SDK 不叠加重试。首选模型连续 3 次没有有效工具调用时，当前步骤改用 `gpt-6-luna`（high），最多再试 3 次；下一步重新使用首选模型。兜底模型及次数可通过 `.env.example` 中的 `OPENAI_TOOL_FALLBACK_*` 和 `ARTEMIS_TOOL_FALLBACK_MAX_NO_PROGRESS_ATTEMPTS` 配置。请求耗尽后沿用任务暂停/恢复机制，鉴权或非法参数立即失败。视频分析与录屏是独立开关，默认均关闭。每项批量用例可覆盖模式、模型、设备、录屏和超时等配置，不影响其他任务。暂停先返回 `pausing`，确认状态为 `task_paused` 才表示设备已释放；Pro 等待模型回复或只读诊断时可及时中断，恢复后重新观察设备。AI 可用 `mobile_manage_task(action="inject_instruction")` 提交指导并恢复；指导由执行器在安全检查点领取，可能中断当前模型等待并重新观察设备；接收回执不保证任务结束前已消费，已经开始的设备动作会安全结束。用 `guidance_history` 核对指导是否被领取、在任务流程中核对实际执行情况，或用 `resume` 单独恢复。
429 等待、兜底模型及次数也可在网页“模型配置 → 稳定性与兜底模型”或通过已授权的 `mobile_config(section="model")` 配置。批次 ZIP 还收录每项任务已保存的步骤截图和录屏。
diff --git a/docs/exploration-mode.md b/docs/exploration-mode.md
index b68bd89..922825e 100644
--- a/docs/exploration-mode.md
+++ b/docs/exploration-mode.md
@@ -8,7 +8,7 @@
## 模型与超时
-探索独立保存模型配置，默认 `gpt-6.1-sol / high`，预设也提供 `gpt-6-astra`，允许自定义模型。执行、人工示范分析、用例提炼和结束后的讨论使用这套配置。修改默认值仅影响新探索；已创建会话保留配置快照，Flash/Pro 默认值不变。
+探索独立保存模型配置，默认 `gpt-6.1-terra / high`，预设也提供 `gpt-6-astra`，允许自定义模型。执行、人工示范分析、用例提炼和结束后的讨论使用这套配置。修改默认值仅影响新探索；已创建会话保留配置快照，Flash/Pro 默认值不变。
默认每段主动执行 1800 秒。模型调用、设备动作、录像分析计入；排队、等待回答和人工示范不计入。到时保存进度并暂停，可点击「继续探索」开启下一段；不会仅因超时把测试判成失败，累计主动耗时仍保留。
diff --git a/mcp_server/README.md b/mcp_server/README.md
index 9fcc3dd..9597660 100644
--- a/mcp_server/README.md
+++ b/mcp_server/README.md
@@ -61,7 +61,7 @@ The bridge discovers a live service from the same installation; `ARTEMIS_DAEMON_
## 🛠️ MCP Tools Overview
-* **`mobile_exploration`**: Guided exploration using the same queue and conversation as the web workspace. Create/start, exchange messages, pause/continue, record and analyze demonstrations, save immutable case versions, rerun with Flash/Pro, or add a batch draft. Writes use stable request IDs; reads never launch work. Defaults are independent (`gpt-6.1-sol / high`) and timeouts pause without marking the test failed. See the [exploration workflow](../docs/exploration-mode.md).
+* **`mobile_exploration`**: Guided exploration using the same queue and conversation as the web workspace. Create/start, exchange messages, pause/continue, record and analyze demonstrations, save immutable case versions, rerun with Flash/Pro, or add a batch draft. Writes use stable request IDs; reads never launch work. Defaults are independent (`gpt-6.1-terra / high`) and timeouts pause without marking the test failed. See the [exploration workflow](../docs/exploration-mode.md).
* **`mobile_run_task`**: Asynchronously launches an autonomous mobile automation task (`Flash` or `Pro` model) with optional `device_serial` targeting. Pro runs can be tuned with `verification_level` (`off` | `final` | `checkpoints` | `strict` — how much the Checker audits) and `explorer_mode` (`flash` | `pro` | `ultra` — the Operator's perception depth).
* **`mobile_manage_task`**: Manages task lifecycle (`status`, `stop`, `inject_instruction`), returning task state and assigned `device_serial`. Pass `release_loop=True` with `inject_instruction` to gracefully end a `[Loop:continuous]` monitoring task — this explicit signal (not "please stop" wording) is what unlocks the loop milestone's completion.
* **`mobile_get_device_state`**: Real-time observer (`screenshot` or OCR+XML `hierarchy`) with optional `device_serial`.
diff --git a/mcp_server/catalog.json b/mcp_server/catalog.json
index 53aedf8..8792315 100644
--- a/mcp_server/catalog.json
+++ b/mcp_server/catalog.json
@@ -437,7 +437,7 @@
},
{
"name": "mobile_config",
- "description": "Read configuration or explicitly save non-secret defaults for FUTURE tasks.\n\nOnly set persist_defaults=true when the user requests saved/global configuration.\nmodel changes: model (preset or custom), base_url, reasoning_effort\n(default/none/minimal/low/medium/high/xhigh/max), timeout_seconds, max_retries,\nretry_interval_seconds, rate_limit_retry_seconds, tool_fallback_model,\ntool_fallback_reasoning_effort, tool_fallback_max_retries,\ntool_fallback_no_progress_attempts. 'default' omits reasoning;\nunsupported values fail visibly.\ncapabilities: video_enabled and recording_enabled are independent; outputter_enabled,\nflash_final_check_enabled (false; Flash exit-only AI review),\nstep_summarizer_enabled, screenshot_limit_enabled, task_timeout_seconds (default 1800),\nobservation_timeout_seconds (60), observation_max_attempts (3),\ndevice_command_timeout_seconds (10), action_failure_attempts (3),\nworker_cleanup_timeout_seconds (30), adb_auto_connect (true),\nadb_discovery_interval_seconds (10), adb_discovery_timeout_seconds (5).\nreports: formats=['docx','xlsx'] (at least one).\nexploration: model (default gpt-6.1-sol), reasoning_effort (high), timeout_seconds\n(1800 active seconds), demonstration_timeout_seconds (900), demonstration_warning_seconds (60).\nAPI keys are never accepted or returned: use the Web UI or .env for credentials.\nSaving settings does not change an already running task. Flash/Pro are execution\nprofiles, not model names. Model retries do not repeat device test cases.\n",
+ "description": "Read configuration or explicitly save non-secret defaults for FUTURE tasks.\n\nOnly set persist_defaults=true when the user requests saved/global configuration.\nmodel changes: model (preset or custom), base_url, reasoning_effort\n(default/none/minimal/low/medium/high/xhigh/max), timeout_seconds, max_retries,\nretry_interval_seconds, rate_limit_retry_seconds, tool_fallback_model,\ntool_fallback_reasoning_effort, tool_fallback_max_retries,\ntool_fallback_no_progress_attempts. 'default' omits reasoning;\nunsupported values fail visibly.\ncapabilities: video_enabled and recording_enabled are independent; outputter_enabled,\nflash_final_check_enabled (false; Flash exit-only AI review),\nstep_summarizer_enabled, screenshot_limit_enabled, task_timeout_seconds (default 1800),\nobservation_timeout_seconds (60), observation_max_attempts (3),\ndevice_command_timeout_seconds (10), action_failure_attempts (3),\nworker_cleanup_timeout_seconds (30), adb_auto_connect (true),\nadb_discovery_interval_seconds (10), adb_discovery_timeout_seconds (5).\nreports: formats=['docx','xlsx'] (at least one).\nexploration: model (default gpt-6.1-terra), reasoning_effort (high), timeout_seconds\n(1800 active seconds), demonstration_timeout_seconds (900), demonstration_warning_seconds (60).\nAPI keys are never accepted or returned: use the Web UI or .env for credentials.\nSaving settings does not change an already running task. Flash/Pro are execution\nprofiles, not model names. Model retries do not repeat device test cases.\n",
"inputSchema": {
"properties": {
"section": {
@@ -791,7 +791,7 @@
},
{
"name": "mobile_exploration",
- "description": "Explore ONE complex test with user guidance and produce a reusable case.\n\ndeletion_preview lists affected conversation runs; delete requires user_confirmed=true\nafter the user requests deletion. Independent validation runs are preserved.\nRead-only list/get/case never execute tasks. create stores {goal,config?,task_config?};\nstart/continue explicitly enqueue or resume Pro exploration. Independent defaults:\ngpt-6.1-sol/high, 1800 active seconds; timeout pauses, never fails by itself.\nmessage payload={text}; answer pending questions or guide without discarding history.\ndemonstrate payload={source:'android'|'computer'|'both'} pauses at a safe point and\nrecords while retaining the device/plugin lease. Wait for recording before the user\noperates. demo_finish analyzes the SAVED clip; demo_extend payload={seconds:900};\ndemo_cancel preserves evidence. Default 900s, warning 60s before expiry, auto-pause.\nsave_case payload={expected_version,case:{name,goal,preconditions?,evidence?}};\ngenerate payload={source_run_id,expected_version} starts evidence-based synthesis,\npoll get.generations for completion. validate payload={version,profile,task_config?}\nexplicitly runs the saved version in Flash/Pro; batch additionally accepts name and\ncreates a DRAFT batch, never starts it. Exploration pass and rerun pass are separate.\nUse stable request_id for every mutation except pause/stop/delete, reusing it on transport\nretries. No API keys; configure secrets in the Web UI. mobile_config(section='exploration')\ncontrols non-secret defaults. After errors call mobile_diagnose before resubmitting.\n",
+ "description": "Explore ONE complex test with user guidance and produce a reusable case.\n\ndeletion_preview lists affected conversation runs; delete requires user_confirmed=true\nafter the user requests deletion. Independent validation runs are preserved.\nRead-only list/get/case never execute tasks. create stores {goal,config?,task_config?};\nstart/continue explicitly enqueue or resume Pro exploration. Independent defaults:\ngpt-6.1-terra/high, 1800 active seconds; timeout pauses, never fails by itself.\nmessage payload={text}; answer pending questions or guide without discarding history.\ndemonstrate payload={source:'android'|'computer'|'both'} pauses at a safe point and\nrecords while retaining the device/plugin lease. Wait for recording before the user\noperates. demo_finish analyzes the SAVED clip; demo_extend payload={seconds:900};\ndemo_cancel preserves evidence. Default 900s, warning 60s before expiry, auto-pause.\nsave_case payload={expected_version,case:{name,goal,preconditions?,evidence?}};\ngenerate payload={source_run_id,expected_version} starts evidence-based synthesis,\npoll get.generations for completion. validate payload={version,profile,task_config?}\nexplicitly runs the saved version in Flash/Pro; batch additionally accepts name and\ncreates a DRAFT batch, never starts it. Exploration pass and rerun pass are separate.\nUse stable request_id for every mutation except pause/stop/delete, reusing it on transport\nretries. No API keys; configure secrets in the Web UI. mobile_config(section='exploration')\ncontrols non-secret defaults. After errors call mobile_diagnose before resubmitting.\n",
"inputSchema": {
"properties": {
"action": {
diff --git a/mcp_server/tools/automation.py b/mcp_server/tools/automation.py
index 08e3032..ae45dae 100644
--- a/mcp_server/tools/automation.py
+++ b/mcp_server/tools/automation.py
@@ -200,7 +200,7 @@ async def mobile_config(
worker_cleanup_timeout_seconds (30), adb_auto_connect (true),
adb_discovery_interval_seconds (10), adb_discovery_timeout_seconds (5).
reports: formats=['docx','xlsx'] (at least one).
- exploration: model (default gpt-6.1-sol), reasoning_effort (high), timeout_seconds
+ exploration: model (default gpt-6.1-terra), reasoning_effort (high), timeout_seconds
(1800 active seconds), demonstration_timeout_seconds (900), demonstration_warning_seconds (60).
API keys are never accepted or returned: use the Web UI or .env for credentials.
Saving settings does not change an already running task. Flash/Pro are execution
diff --git a/mcp_server/tools/exploration.py b/mcp_server/tools/exploration.py
index 2a12bd7..8414b61 100644
--- a/mcp_server/tools/exploration.py
+++ b/mcp_server/tools/exploration.py
@@ -43,7 +43,7 @@ async def mobile_exploration(
after the user requests deletion. Independent validation runs are preserved.
Read-only list/get/case never execute tasks. create stores {goal,config?,task_config?};
start/continue explicitly enqueue or resume Pro exploration. Independent defaults:
- gpt-6.1-sol/high, 1800 active seconds; timeout pauses, never fails by itself.
+ gpt-6.1-terra/high, 1800 active seconds; timeout pauses, never fails by itself.
message payload={text}; answer pending questions or guide without discarding history.
demonstrate payload={source:'android'|'computer'|'both'} pauses at a safe point and
records while retaining the device/plugin lease. Wait for recording before the user
diff --git a/skills/artemis-use/references/exploration.md b/skills/artemis-use/references/exploration.md
index c1b7b3e..bb27ea3 100644
--- a/skills/artemis-use/references/exploration.md
+++ b/skills/artemis-use/references/exploration.md
@@ -6,7 +6,7 @@
## 模型与超时
-探索独立保存模型配置，默认 `gpt-6.1-sol / high`，预设也提供 `gpt-6-astra`，允许自定义模型。执行、人工示范分析、用例提炼和结束后的讨论使用这套配置。修改默认值仅影响新探索；已创建会话保留配置快照，Flash/Pro 默认值不变。
+探索独立保存模型配置，默认 `gpt-6.1-terra / high`，预设也提供 `gpt-6-astra`，允许自定义模型。执行、人工示范分析、用例提炼和结束后的讨论使用这套配置。修改默认值仅影响新探索；已创建会话保留配置快照，Flash/Pro 默认值不变。
默认每段主动执行 1800 秒。模型调用、设备动作、录像分析计入；排队、等待回答和人工示范不计入。到时保存进度并暂停，可点击「继续探索」开启下一段；不会仅因超时把测试判成失败，累计主动耗时仍保留。
diff --git a/skills/artemis-use/references/modes.md b/skills/artemis-use/references/modes.md
index 3532ab7..f5a261e 100644
--- a/skills/artemis-use/references/modes.md
+++ b/skills/artemis-use/references/modes.md
@@ -17,7 +17,7 @@
## 模式、模型与配置
- `mobile_run_task.model` 取 `Flash` / `Pro`。API 模型名放 `runtime_overrides.model`；思考强度放 `runtime_overrides.reasoning_effort`。批次使用 `task_config.profile` 选择模式，`task_config.model` 选择 API 模型。
-- 探索独立配置，默认 `gpt-6.1-sol/high`，支持 `gpt-6-astra` 及已配置网关支持的自定义模型。执行、示范分析、提炼共用该探索配置；不改变 Flash/Pro 默认值。
+- 探索独立配置，默认 `gpt-6.1-terra/high`，支持 `gpt-6-astra` 及已配置网关支持的自定义模型。执行、示范分析、提炼共用该探索配置；不改变 Flash/Pro 默认值。
- 默认 high、单次模型请求时限 300 秒。`default` 表示不发送思考强度，`none` 是显式关闭，不能混为一谈；接口不支持参数时如实报告，不擅自降级。
- 请求时限与整任务主动执行时限不同。排队/用户暂停不计入任务主动执行；探索等待回答、人工示范也不计入。
- 录屏 `record_video` 和 AI 视频分析 `analyze_video` 独立，默认关闭。需要可下载录像时明确启用录屏。
diff --git a/tests/unit/admin_console/test_model_settings.py b/tests/unit/admin_console/test_model_settings.py
index cecf8d6..f35a384 100644
--- a/tests/unit/admin_console/test_model_settings.py
+++ b/tests/unit/admin_console/test_model_settings.py
@@ -96,6 +96,7 @@ def test_model_presets_keep_supported_choices_without_restricting_custom_names()
presets = service.get_model_settings()["model_presets"]
assert presets == [
"gpt-6-luna",
+ "gpt-6.1-terra",
"gpt-6.1-sol",
"gpt-6-astra",
"mimo-v2.6-pro",
diff --git a/tests/unit/test_exploration_service.py b/tests/unit/test_exploration_service.py
index 4aeab29..80d4cca 100644
--- a/tests/unit/test_exploration_service.py
+++ b/tests/unit/test_exploration_service.py
@@ -30,7 +30,7 @@ async def test_admission_uses_shared_queue_with_independent_model(service, monke
await service.start(session["id"], "start")
request = enqueue.call_args.args[0]
assert request.profile == "pro"
- assert request.runtime_overrides.model == "gpt-6.1-sol"
+ assert request.runtime_overrides.model == "gpt-6.1-terra"
assert request.runtime_overrides.reasoning_effort == "high"
assert request.runtime_overrides.verification_level == "final"
assert request.conversation_id == session["id"]
@@ -193,7 +193,7 @@ def test_independent_saved_defaults_do_not_change_existing_sessions(service):
session = service.create(CreateExploration(goal="探索设置", request_id="create"))
service.save_settings(ExplorationConfig(model="gpt-6-astra", reasoning_effort="max"))
later = service.create(CreateExploration(goal="另一次探索", request_id="later"))
- assert service.repository.get(session["id"])["config"]["model"] == "gpt-6.1-sol"
+ assert service.repository.get(session["id"])["config"]["model"] == "gpt-6.1-terra"
assert later["config"]["model"] == "gpt-6-astra"
diff --git a/tests/unit/test_exploration_store.py b/tests/unit/test_exploration_store.py
index e8b0502..2bf2919 100644
--- a/tests/unit/test_exploration_store.py
+++ b/tests/unit/test_exploration_store.py
@@ -187,7 +187,7 @@ def test_generated_case_cannot_borrow_another_sessions_evidence(repository):
def test_exploration_config_defaults_and_rejects_credentials():
config = ExplorationConfig()
assert (config.model, config.reasoning_effort, config.timeout_seconds) == (
- "gpt-6.1-sol",
+ "gpt-6.1-terra",
"high",
1800,
)