auto-coder 0.1.254__py3-none-any.whl → 0.1.255__py3-none-any.whl

{auto_coder-0.1.254.dist-info → auto_coder-0.1.255.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: auto-coder
-Version: 0.1.254
+Version: 0.1.255
 Summary: AutoCoder: AutoCoder
 Author: allwefantasy
 Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence

{auto_coder-0.1.254.dist-info → auto_coder-0.1.255.dist-info}/RECORD

@@ -1,17 +1,17 @@
 autocoder/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-autocoder/auto_coder.py,sha256=KTc-twhc4MNabzgNWj8nDZZTD-uMneDABnFSWnt6Ta8,64640
+autocoder/auto_coder.py,sha256=uzNAtguu7O2m7vve1H2FQnWCmVuqT07RITxS5jgWz4U,64774
 autocoder/auto_coder_lang.py,sha256=Rtupq6N3_HT7JRhDKdgCBcwRaiAnyCOR_Gsp4jUomrI,3229
 autocoder/auto_coder_rag.py,sha256=DDAmqw36CO6phtdQuN8LYIbIR3YGdoZw5-pG0LjVxMc,29063
 autocoder/auto_coder_rag_client_mcp.py,sha256=WV7j5JUiQge0x4-B7Hp5-pSAFXLbvLpzQMcCovbauIM,6276
 autocoder/auto_coder_rag_mcp.py,sha256=-RrjNwFaS2e5v8XDIrKR-zlUNUE8UBaeOtojffBrvJo,8521
 autocoder/auto_coder_server.py,sha256=XU9b4SBH7zjPPXaTWWHV4_zJm-XYa6njuLQaplYJH_c,20290
 autocoder/benchmark.py,sha256=Ypomkdzd1T3GE6dRICY3Hj547dZ6_inqJbBJIp5QMco,4423
-autocoder/chat_auto_coder.py,sha256=1QgTy6SiZjN2mkYe5pVvPTdZyedwijDctTUS8tSukVM,107527
-autocoder/chat_auto_coder_lang.py,sha256=OiwjnOqbiWwbaVuOhehEqdUFPOJbhfvQr3sIq3b0xj4,17455
+autocoder/chat_auto_coder.py,sha256=a1YEp6OPMzpLbRpr2hrbzF6pRnhVPTVxyZfBiQHFPIw,109283
+autocoder/chat_auto_coder_lang.py,sha256=1cJrjFGrcOQnuP2LdZpgGDSX4CNaIYI7KZGvEEtj6_Q,18242
 autocoder/command_args.py,sha256=9aYJ-AmPxP1sQh6ciw04FWHjSn31f2W9afXFwo8wgx4,30441
 autocoder/lang.py,sha256=U6AjVV8Rs1uLyjFCZ8sT6WWuNUxMBqkXXIOs4S120uk,14511
-autocoder/models.py,sha256=uo4mFWb-kdtd-8e2HPfCuXjhkDsE37n1dY6Nt1SU6kk,7625
-autocoder/version.py,sha256=LLYeCF3eLt_aBS3HsiF4z5DiLpgabod5UhddNdga9mk,23
+autocoder/models.py,sha256=0f653gjpQN_JO5k7h6wmTF4bVd6CW3fpQOyHIZ3ZUv4,7558
+autocoder/version.py,sha256=vxd_TSbCxiXnBHQnfprB361BOJEhjQMPeDRseFI3YrY,23
 autocoder/agent/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/agent/auto_demand_organizer.py,sha256=NWSAEsEk94vT3lGjfo25kKLMwYdPcpy9e-i21txPasQ,6942
 autocoder/agent/auto_filegroup.py,sha256=CW7bqp0FW1GIEMnl-blyAc2UGT7O9Mom0q66ITz1ckM,6635
@@ -29,7 +29,7 @@ autocoder/common/__init__.py,sha256=6maackdzrYnUPvpgVPl92JdMOnw7X4n3EnEQA9OnLGE,
 autocoder/common/anything2images.py,sha256=0ILBbWzY02M-CiWB-vzuomb_J1hVdxRcenAfIrAXq9M,25283
 autocoder/common/anything2img.py,sha256=4TREa-sOA-iargieUy7MpyCYVUE-9Mmq0wJtwomPqnE,7662
 autocoder/common/audio.py,sha256=Kn9nWKQddWnUrAz0a_ZUgjcu4VUU_IcZBigT7n3N3qc,7439
-autocoder/common/auto_coder_lang.py,sha256=9qaluOOt0X6soURfQs27_jw2tIA95yoLdAcACt-Zsf4,16359
+autocoder/common/auto_coder_lang.py,sha256=nF8XrHpSbibk6ro8Oum-0V0FXcDS1lHP_hL-CnPJtT4,16974
 autocoder/common/buildin_tokenizer.py,sha256=L7d5t39ZFvUd6EoMPXUhYK1toD0FHlRH1jtjKRGokWU,1236
 autocoder/common/chunk_validation.py,sha256=BrR_ZWavW8IANuueEE7hS8NFAwEvm8TX34WnPx_1hs8,3030
 autocoder/common/cleaner.py,sha256=NU72i8C6o9m0vXExab7nao5bstBUsfJFcj11cXa9l4U,1089
@@ -43,7 +43,7 @@ autocoder/common/code_auto_merge_diff.py,sha256=qpEuHJEgX6sWK7EDFEKqcYkyI28wOyM4
 autocoder/common/code_auto_merge_editblock.py,sha256=sxgYMLMACRwJvw-bABkdDHezPelsDFrOCpGuhtT5Dzs,17504
 autocoder/common/code_auto_merge_strict_diff.py,sha256=P0nKNkBrFMybTSZ7kOdA_JixoVmLCZIhAP5q7ILJ9j0,9538
 autocoder/common/code_modification_ranker.py,sha256=qfadP9P-iiidCG2A_MjAf3Ca8cMz7YlnN08D_kH6uFc,6447
-autocoder/common/command_completer.py,sha256=SSeb8MDH0JPvfdyW-S2uaHnui4VBDfSQvQPLbv3ORPA,9314
+autocoder/common/command_completer.py,sha256=IShrZJSpR-Q_MCj_aCVdVyscLYDKj5ZQK357QBcQ_oQ,9420
 autocoder/common/command_generator.py,sha256=-hmbD_AnCa5HxL4BznuEfYAf_l8AxU5fAG5F0sM_fuE,2116
 autocoder/common/command_templates.py,sha256=mnB3n8i0yjH1mqzyClEg8Wpr9VbZV44kxky66Zu6OJY,8557
 autocoder/common/const.py,sha256=eTjhjh4Aj4CUzviJ81jaf3Y5cwqsLATySn2wJxaS6RQ,2911
@@ -57,6 +57,7 @@ autocoder/common/mcp_hub.py,sha256=2ZyJv3Aiv4Y97UHut49oYhIFcu7ICR-mptDEBSgT3uE,1
 autocoder/common/mcp_server.py,sha256=1G6e0IbeS_h7CA1vr0dPAnf0o2H1f1A8I4bua8EUtKw,12318
 autocoder/common/mcp_tools.py,sha256=KsLvRrB6pvmebqd-lDaSH6IBJR0AIxWRE-dtCEG_w9k,12485
 autocoder/common/memory_manager.py,sha256=2ZjYG7BPyvbYalZBF6AM_G5e10Qkw_zrqtD4Zd7GSsQ,3663
+autocoder/common/model_speed_test.py,sha256=U48xUUpOnbwUal1cdij4YAn_H2PD2pNaqrMHaYtQRfI,15200
 autocoder/common/printer.py,sha256=P1WU0QjlfnjqTP5uA55GkHZCpFzRPFkc34DMMandreg,2023
 autocoder/common/recall_validation.py,sha256=Avt9Q9dX3kG6Pf2zsdlOHmsjd-OeSj7U1PFBDp_Cve0,1700
 autocoder/common/screenshots.py,sha256=_gA-z1HxGjPShBrtgkdideq58MG6rqFB2qMUJKjrycs,3769
@@ -69,6 +70,7 @@ autocoder/common/types.py,sha256=PXTETrsTvhLE49jqAeUKGySvxBN9pjeyCgRHLDYdd9U,664
 autocoder/common/utils_code_auto_generate.py,sha256=kDW5B_2wRLk7hAls2hewliDacV86lrPz8Jan01BvtCw,3573
 autocoder/common/mcp_servers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/common/mcp_servers/mcp_server_perplexity.py,sha256=IXTyMpd1CQcBLzVinA-_OIOHoNmbzvuW6pXIadaKHJE,5533
+autocoder/data/byzerllm.md,sha256=SGCMpEaUQ0ysPxQsgzyyp5sgvEr8dZsxEGAfVcPBIq0,47741
 autocoder/data/tokenizer.json,sha256=7Lb5_DaYlDRvBRH0B0ynXO5c1fOwbQLxujX805-OEh0,7847602
 autocoder/db/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/db/store.py,sha256=tFT66bP2ZKIqZip-uhLkHRSLaaOAUUDZfozJwcqix3c,1908
@@ -144,9 +146,9 @@ autocoder/utils/types.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/utils/auto_coder_utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/utils/auto_coder_utils/chat_stream_out.py,sha256=xWXqICANbDOovH4wcFW1eSI7lB7TjXbk1mSU4bTKEW4,11434
 autocoder/utils/chat_auto_coder_utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-auto_coder-0.1.254.dist-info/LICENSE,sha256=HrhfyXIkWY2tGFK11kg7vPCqhgh5DcxleloqdhrpyMY,11558
-auto_coder-0.1.254.dist-info/METADATA,sha256=a7eZ5YDrFy4_RSOLcTBzbdye35TnHukamLS-g3-jNkM,2616
-auto_coder-0.1.254.dist-info/WHEEL,sha256=GV9aMThwP_4oNCtvEC2ec3qUYutgWeAzklro_0m4WJQ,91
-auto_coder-0.1.254.dist-info/entry_points.txt,sha256=0nzHtHH4pNcM7xq4EBA2toS28Qelrvcbrr59GqD_0Ak,350
-auto_coder-0.1.254.dist-info/top_level.txt,sha256=Jqc0_uJSw2GwoFQAa9iJxYns-2mWla-9ok_Y3Gcznjk,10
-auto_coder-0.1.254.dist-info/RECORD,,
+auto_coder-0.1.255.dist-info/LICENSE,sha256=HrhfyXIkWY2tGFK11kg7vPCqhgh5DcxleloqdhrpyMY,11558
+auto_coder-0.1.255.dist-info/METADATA,sha256=2uB08jgGHyp3_DWMI2_vxoFoptVJO76Va-yek1umBac,2616
+auto_coder-0.1.255.dist-info/WHEEL,sha256=GV9aMThwP_4oNCtvEC2ec3qUYutgWeAzklro_0m4WJQ,91
+auto_coder-0.1.255.dist-info/entry_points.txt,sha256=0nzHtHH4pNcM7xq4EBA2toS28Qelrvcbrr59GqD_0Ak,350
+auto_coder-0.1.255.dist-info/top_level.txt,sha256=Jqc0_uJSw2GwoFQAa9iJxYns-2mWla-9ok_Y3Gcznjk,10
+auto_coder-0.1.255.dist-info/RECORD,,

autocoder/auto_coder.py

@@ -256,11 +256,13 @@ def main(input_args: Optional[List[str]] = None):
             libs_dir = os.path.join(auto_coder_dir, "storage", "libs")
             code_search_path = None
             if os.path.exists(libs_dir):
-                retrieval_libs_dir = os.path.join(
-                    libs_dir, get_latest_byzer_retrieval_lib(libs_dir)
-                )
-                if os.path.exists(retrieval_libs_dir):
-                    code_search_path = [retrieval_libs_dir]
+                latest_retrieval_lib_dir = get_latest_byzer_retrieval_lib(libs_dir)
+                if latest_retrieval_lib_dir :
+                    retrieval_libs_dir = os.path.join(
+                        libs_dir, latest_retrieval_lib_dir
+                    )
+                    if os.path.exists(retrieval_libs_dir):
+                        code_search_path = [retrieval_libs_dir]
 
             try:
                 init_options = {}

autocoder/chat_auto_coder.py

@@ -2197,24 +2197,49 @@ def manage_models(params, query: str):
         subcmd = "/remove"
         query = query.replace("/remove", "", 1).strip()
 
+    if "/speed-test" in query:
+        subcmd = "/speed-test"
+        query = query.replace("/speed-test", "", 1).strip()
+
+    if "/speed_test" in query:
+        subcmd = "/speed-test"
+        query = query.replace("/speed_test", "", 1).strip() 
+
+    if "input_price" in query:
+        subcmd = "/input_price"
+        query = query.replace("/input_price", "", 1).strip()
+
+    if "output_price" in query:
+        subcmd = "/output_price"
+        query = query.replace("/output_price", "", 1).strip()        
+            
+    if "/speed" in query:
+        subcmd = "/speed"
+        query = query.replace("/speed", "", 1).strip()
+    
+        
+
     if not subcmd:
-        printer.print_in_terminal("models_usage")
-        return    
+        printer.print_in_terminal("models_usage")        
 
     if subcmd == "/list":                    
         if models_data:
+            # Sort models by speed (average_speed)
+            sorted_models = sorted(models_data, key=lambda x: float(x.get('average_speed', 0)))
+            sorted_models.reverse()
+            
             table = Table(
                 title=printer.get_message_from_key("models_title"),
                 expand=True,
                 show_lines=True
             )
-            table.add_column("Name", style="cyan", width=40, no_wrap=False)
-            table.add_column("Model Name", style="magenta", width=30, overflow="fold")
-            table.add_column("Base URL", style="white", width=50, overflow="fold")
-            table.add_column("Input Price (M)", style="magenta", width=15)
-            table.add_column("Output Price (M)", style="magenta", width=15)
-            table.add_column("Speed (s/req)", style="blue", width=15)
-            for m in models_data:
+            table.add_column("Name", style="cyan", width=30, overflow="fold", no_wrap=False)
+            table.add_column("Model Name", style="magenta", width=30, overflow="fold", no_wrap=False)
+            table.add_column("Base URL", style="white", width=40, overflow="fold", no_wrap=False)
+            table.add_column("Input Price (M)", style="magenta", width=15, overflow="fold", no_wrap=False)
+            table.add_column("Output Price (M)", style="magenta", width=15, overflow="fold", no_wrap=False)
+            table.add_column("Speed (s/req)", style="blue", width=15, overflow="fold", no_wrap=False)
+            for m in sorted_models:
                 # Check if api_key_path exists and file exists
                 is_api_key_set = "api_key" in m  
                 name = m.get("name", "")              
@@ -2281,6 +2306,26 @@ def manage_models(params, query: str):
         else:
             printer.print_in_terminal("models_speed_usage", style="red")
             
+    elif subcmd == "/speed-test":
+        from autocoder.common.model_speed_test import render_speed_test_in_terminal
+        test_rounds = 1  # 默认测试轮数
+        
+        enable_long_context = False
+        if "/long_context" in query:
+            enable_long_context = True
+            query = query.replace("/long_context", "", 1).strip()
+
+        if "/long-context" in query:
+            enable_long_context = True
+            query = query.replace("/long-context", "", 1).strip()
+
+        # 解析可选的测试轮数参数
+        args = query.strip().split()
+        if args and args[0].isdigit():
+            test_rounds = int(args[0])
+            
+        render_speed_test_in_terminal(params.product_mode, test_rounds,enable_long_context=enable_long_context)
+    
     elif subcmd == "/add":
         # Support both simplified and legacy formats
         args = query.strip().split(" ")        

autocoder/chat_auto_coder_lang.py

@@ -85,7 +85,7 @@ MESSAGES = {
         "design_desc": "Generate SVG image based on the provided description",
         "commit_desc": "Auto generate yaml file and commit changes based on user's manual changes",
         "models_desc": "Manage model configurations, only available in lite mode",
-        "models_usage": "Usage: /models /list|/add|/add_model|/remove|/price|/speed ...",
+        "models_usage": "Usage: /models <command>\nAvailable subcommands:\n  /list - List all models\n  /add <name> <api_key> - Add a built-in model\n  /add_model - Add a custom model\n  /remove <name> - Remove a model\n  /input_price <name> <value> - Set model input price\n  /output_price <name> <value> - Set model output price\n  /speed <name> <value> - Set model speed\n  /speed-test - Test models speed\n  /speed-test-long - Test models speed with long context",
         "models_added": "Added/Updated model '{{name}}' successfully.",
         "models_add_failed": "Failed to add model '{{name}}'. Model not found in defaults.",
         "models_add_usage": "Usage: /models /add <name> <api_key> or\n/models /add <name> <model_type> <model_name> <base_url> <api_key_path> [description]",
@@ -213,7 +213,7 @@ MESSAGES = {
         "conf_value": "值",
         "conf_title": "配置设置",
         "conf_subtitle": "使用 /conf <key>:<value> 修改这些设置",
-        "models_usage": "用法: /models /list|/add|/add_model|/remove|/price|/speed ...",
+        "models_usage": "用法: /models <命令>\n可用的子命令:\n  /list - 列出所有模型\n  /add <名称> <API密钥> - 添加内置模型\n  /add_model - 添加自定义模型\n  /remove <名称> - 移除模型\n  /input_price <名称> <价格> - 设置模型输入价格\n  /output_price <名称> <价格> - 设置模型输出价格\n  /speed <名称> <速度> - 设置模型速度\n  /speed-test - 测试模型速度\n  /speed-test-long - 使用长文本上下文测试模型速度",
         "models_added": "成功添加/更新模型 '{{name}}'。",
         "models_add_failed": "添加模型 '{{name}}' 失败。在默认模型中未找到该模型。",
         "models_add_usage": "用法: /models /add <name> <api_key> 或\n/models /add <name> <model_type> <model_name> <base_url> <api_key_path> [description]",

autocoder/common/auto_coder_lang.py

@@ -3,6 +3,11 @@ from byzerllm.utils import format_str_jinja2
 
 MESSAGES = {
     "en": {
+        "models_no_active": "No active models found",
+        "models_speed_test_results": "Model Speed Test Results",
+        "models_testing": "Testing model: {{name}}...",
+        "models_testing_start": "Starting speed test for all active models...",
+        "models_testing_progress": "Testing progress: {{ completed }}/{{ total }} models",
         "generation_cancelled": "[Interrupted] Generation cancelled",
         "model_not_found": "Model {{model_name}} not found",
         "generating_shell_script": "Generating Shell Script",
@@ -92,6 +97,10 @@ MESSAGES = {
         "estimated_input_tokens_in_generate": "Estimated input tokens in generate ({{ generate_mode }}): {{ estimated_input_tokens }}",
     },
     "zh": {
+        "models_no_active": "未找到激活的模型",
+        "models_speed_test_results": "模型速度测试结果",
+        "models_testing": "正在测试模型: {{name}}...",
+        "models_testing_start": "开始对所有激活的模型进行速度测试...",
         "generation_cancelled": "[已中断] 生成已取消",
         "model_not_found": "未找到模型: {{model_name}}",
         "generating_shell_script": "正在生成 Shell 脚本",

autocoder/common/command_completer.py

@@ -33,7 +33,11 @@ COMMANDS = {
         "/add": "",
         "/add_model": "",
         "/remove": "",
-        "/list": ""        
+        "/list": "",
+        "/speed": "",
+        "/speed-test": "",
+        "/input_price": "",
+        "/output_price": "",       
     }
 }
 

autocoder/common/model_speed_test.py

@@ -0,0 +1,392 @@
+import time
+import byzerllm
+from typing import Dict, Any, List, Optional
+from rich.console import Console
+from rich.table import Table
+from rich.panel import Panel
+from autocoder.common.printer import Printer
+from autocoder import models as models_module
+from autocoder.utils.llms import get_single_llm
+import byzerllm
+import pkg_resources
+from concurrent.futures import ThreadPoolExecutor
+from typing import Dict, List, Tuple
+from pydantic import BaseModel
+
+class ModelSpeedTestResult(BaseModel):
+    model_name: str
+    tokens_per_second: float
+    first_token_time: float
+    input_tokens_count: float
+    generated_tokens_count: float
+    input_tokens_cost: float
+    generated_tokens_cost: float
+    status: str
+    error: Optional[str] = None
+
+class SpeedTestResults(BaseModel):
+    results: List[ModelSpeedTestResult]
+
+byzerllm_content = ""
+try:
+    byzerllm_conten_path = pkg_resources.resource_filename(
+        "autocoder", "data/byzerllm.md"
+    )
+    with open(byzerllm_conten_path, "r",encoding="utf-8") as f:
+        byzerllm_content = f.read()
+except FileNotFoundError:
+    pass
+
+@byzerllm.prompt()
+def long_context_prompt() -> str:
+    '''
+    下面是我们提供的一份文档：
+    <document>
+    {{ content }}
+    </document>
+    
+    请根据上述文档，实现用户的需求：
+
+    <query>
+    我想开发一个翻译程序，使用prompt 函数实现。
+    </query>
+    '''
+    return {
+        "content": byzerllm_content
+    }
+
+@byzerllm.prompt()
+def short_context_prompt() -> str:
+    '''
+    Hello, can you help me test the response speed?
+    '''
+    return {}
+
+def test_model_speed(model_name: str, 
+                    product_mode: str, 
+                    test_rounds: int = 3,
+                    enable_long_context: bool = False
+                    ) -> Dict[str, Any]:
+    from autocoder.models import get_model_by_name
+    """
+    测试单个模型的速度
+    
+    Args:
+        model_name: 模型名称
+        product_mode: 产品模式 (lite/pro)
+        test_rounds: 测试轮数
+        
+    Returns:
+        Dict包含测试结果:
+            - avg_time: 平均响应时间
+            - min_time: 最小响应时间
+            - max_time: 最大响应时间
+            - first_token_time: 首token时间
+            - success: 是否测试成功
+            - error: 错误信息(如果有)
+    """
+    try:
+        llm = get_single_llm(model_name, product_mode)
+        model_info = get_model_by_name(model_name)
+
+        times = []
+        first_token_times = []
+        tokens_per_seconds = []
+        input_tokens_counts = []
+        generated_tokens_counts = []
+        
+        input_tokens_costs = []
+        generated_tokens_costs = []
+
+        input_tokens_cost_per_m = model_info.get("input_price", 0.0) / 1000000
+        output_tokens_cost_per_m = model_info.get("output_price", 0.0) / 1000000
+
+        test_query = short_context_prompt.prompt()
+        if enable_long_context:
+            test_query = long_context_prompt.prompt()                
+        
+        content = ""
+        for _ in range(test_rounds):
+            start_time = time.time()
+            first_token_received = False
+            first_token_time = None
+            last_meta = None
+            input_tokens_count = 0
+            generated_tokens_count = 0 
+            input_tokens_cost = 0
+            generated_tokens_cost = 0           
+            for chunk,meta in llm.stream_chat_oai(conversations=[{
+                "role": "user",
+                "content": test_query
+            }],delta_mode=True):
+                content += chunk
+                last_meta = meta                
+                current_time = time.time()
+                if not first_token_received:
+                    first_token_time = current_time - start_time
+                    first_token_received = True
+                    first_token_times.append(first_token_time)
+            
+            end_time = time.time()            
+            generated_tokens_count = 0
+            if last_meta:                
+                generated_tokens_count = last_meta.generated_tokens_count
+                input_tokens_count = last_meta.input_tokens_count
+                input_tokens_cost = input_tokens_count * input_tokens_cost_per_m
+                generated_tokens_cost = generated_tokens_count * output_tokens_cost_per_m
+
+                input_tokens_costs.append(input_tokens_cost)
+                generated_tokens_costs.append(generated_tokens_cost)
+                generated_tokens_counts.append(generated_tokens_count)
+                input_tokens_counts.append(input_tokens_count)
+            
+            tokens_per_seconds.append(generated_tokens_count / (end_time - start_time))    
+            times.append(end_time - start_time)
+            
+            
+        avg_time = sum(times) / len(times)
+        return {
+            "tokens_per_second": sum(tokens_per_seconds) / len(tokens_per_seconds),
+            "avg_time": avg_time,
+            "min_time": min(times),
+            "max_time": max(times),
+            "first_token_time": sum(first_token_times) / len(first_token_times),
+            "input_tokens_count": sum(input_tokens_counts) / len(input_tokens_counts),
+            "generated_tokens_count": sum(generated_tokens_counts) / len(generated_tokens_counts),
+            "success": True,
+            "error": None,
+            "input_tokens_cost": sum(input_tokens_costs) / len(input_tokens_costs),
+            "generated_tokens_cost": sum(generated_tokens_costs) / len(generated_tokens_costs)
+        }
+    except Exception as e:
+        return {
+            "tokens_per_second": 0,
+            "avg_time": 0,
+            "min_time": 0,
+            "max_time": 0,
+            "first_token_time": 0,
+            "input_tokens_count": 0,
+            "generated_tokens_count": 0,
+            "success": False,
+            "error": str(e),
+            "input_tokens_cost": 0.0,
+            "generated_tokens_cost": 0.0
+        }
+
+def test_model_speed_wrapper(args: Tuple[str, str, int, bool]) -> Tuple[str, Dict[str, Any]]:
+    """
+    包装测试函数以适应线程池调用
+    
+    Args:
+        args: (model_name, product_mode, test_rounds)的元组
+        
+    Returns:
+        (model_name, test_results)的元组
+    """
+    model_name, product_mode, test_rounds,enable_long_context = args
+    results = test_model_speed(model_name, product_mode, test_rounds,enable_long_context)
+    return (model_name, results)
+
+
+def run_speed_test(product_mode: str, test_rounds: int = 3, max_workers: Optional[int] = None, enable_long_context: bool = False) -> SpeedTestResults:
+    """
+    运行所有已激活模型的速度测试
+    
+    Args:
+        product_mode: 产品模式 (lite/pro)
+        test_rounds: 每个模型测试的轮数
+        max_workers: 最大线程数,默认为None(ThreadPoolExecutor会自动设置)
+        enable_long_context: 是否启用长文本上下文测试
+    
+    Returns:
+        SpeedTestResults: 包含所有模型测试结果的pydantic模型
+    """
+    # 获取所有模型
+    models_data = models_module.load_models()
+    active_models = [m for m in models_data if "api_key" in m] if product_mode == "lite" else models_data
+    
+    if not active_models:
+        return SpeedTestResults(results=[])
+    
+    # 准备测试参数
+    test_args = [(model["name"], product_mode, test_rounds, enable_long_context) for model in active_models]
+    
+    # 存储结果用于排序
+    results_list = []
+    
+    # 使用线程池并发测试
+    with ThreadPoolExecutor(max_workers=max_workers) as executor:
+        # 提交所有测试任务并获取future对象
+        future_to_model = {executor.submit(test_model_speed_wrapper, args): args[0] 
+                          for args in test_args}
+        
+        # 收集结果
+        for future in future_to_model:
+            model_name = future_to_model[future]
+            
+            try:
+                _, results = future.result()
+                
+                if results["success"]:
+                    status = "✓"
+                    results_list.append((
+                        results['tokens_per_second'],
+                        ModelSpeedTestResult(
+                            model_name=model_name,
+                            tokens_per_second=results['tokens_per_second'],
+                            first_token_time=results['first_token_time'],
+                            input_tokens_count=results['input_tokens_count'],
+                            generated_tokens_count=results['generated_tokens_count'],
+                            status=status,
+                            input_tokens_cost=results['input_tokens_cost'],
+                            generated_tokens_cost=results['generated_tokens_cost'],                            
+                        )
+                    ))
+                    try:
+                        # 更新模型的平均速度
+                        models_module.update_model_speed(model_name, results['tokens_per_second'])
+                    except Exception:
+                        pass
+                else:                    
+                    results_list.append((
+                        0,
+                        ModelSpeedTestResult(
+                            model_name=model_name,
+                            tokens_per_second=0,
+                            first_token_time=0,
+                            input_tokens_count=0,
+                            generated_tokens_count=0,
+                            status=f"✗ {results['error']}",
+                            error=results['error'],
+                            input_tokens_cost=0.0,
+                            generated_tokens_cost=0.0
+                        )
+                    ))
+            except Exception as e:
+                results_list.append((
+                    0,
+                    ModelSpeedTestResult(
+                        model_name=model_name,
+                        tokens_per_second=0,
+                        first_token_time=0,
+                        input_tokens_count=0,
+                        generated_tokens_count=0,
+                        status=f"✗ {str(e)}",
+                        error=str(e),
+                        input_tokens_cost=0.0,
+                        generated_tokens_cost=0.0
+                    )
+                ))
+
+    # 按速度排序
+    results_list.sort(key=lambda x: x[0], reverse=True)
+    
+    return SpeedTestResults(results=[result[1] for result in results_list])
+
+def render_speed_test_in_terminal(product_mode: str, test_rounds: int = 3, max_workers: Optional[int] = None,enable_long_context: bool = False) -> None:
+    """
+    运行所有已激活模型的速度测试
+    
+    Args:
+        product_mode: 产品模式 (lite/pro)
+        test_rounds: 每个模型测试的轮数
+        max_workers: 最大线程数,默认为None(ThreadPoolExecutor会自动设置)
+    """
+    printer = Printer()
+    console = Console()
+    
+    # 获取所有模型
+    models_data = models_module.load_models()
+    active_models = [m for m in models_data if "api_key" in m] if product_mode == "lite" else models_data
+    
+    if not active_models:
+        printer.print_in_terminal("models_no_active", style="yellow")
+        return
+        
+    # 创建结果表格
+    table = Table(
+        title=printer.get_message_from_key("models_speed_test_results"),
+        show_header=True,
+        header_style="bold magenta",
+        show_lines=True
+    )
+    
+    table.add_column("Model", style="cyan", width=30)
+    table.add_column("Tokens/s", style="green", width=15)
+    table.add_column("First Token(s)", style="magenta", width=15)
+    table.add_column("Input Tokens", style="magenta", width=15)
+    table.add_column("Generated Tokens", style="magenta", width=15)            
+    table.add_column("Input Tokens Cost", style="yellow", width=15)
+    table.add_column("Generated Tokens Cost", style="yellow", width=15)
+    table.add_column("Status", style="red", width=20)
+    
+    # 准备测试参数
+    test_args = [(model["name"], product_mode, test_rounds, enable_long_context) for model in active_models]
+    
+    # 存储结果用于排序
+    results_list = []
+    
+    # 使用线程池并发测试
+    with ThreadPoolExecutor(max_workers=max_workers) as executor:
+        printer.print_in_terminal("models_testing_start", style="yellow")
+        
+        # 提交所有测试任务并获取future对象
+        future_to_model = {executor.submit(test_model_speed_wrapper, args): args[0] 
+                          for args in test_args}
+        
+        # 收集结果
+        completed = 0
+        total = len(future_to_model)
+        for future in future_to_model:
+            completed += 1
+            printer.print_in_terminal("models_testing_progress", style="yellow", completed=completed, total=total)
+            model_name = future_to_model[future]
+            printer.print_in_terminal("models_testing", style="yellow", name=model_name)
+            
+            try:
+                _, results = future.result()
+                
+                if results["success"]:
+                    status = "✓"
+                    results['status'] = status
+                    results_list.append((
+                        results['tokens_per_second'],
+                        model_name,
+                        results
+                    ))                  
+                    try:  
+                        # 更新模型的平均速度
+                        models_module.update_model_speed(model_name, results['tokens_per_second'])
+                    except Exception as e:
+                        pass
+                else:
+                    status = f"✗ ({results['error']})"
+                    results_list.append((
+                        0,
+                        model_name,
+                        {"tokens_per_second":0,"avg_time": 0, "input_tokens_count":0, "generated_tokens_count":0, "min_time": 0, "max_time": 0, "first_token_time": 0, "input_tokens_cost": 0.0, "generated_tokens_cost": 0.0, "status": status}
+                    ))
+            except Exception as e:
+                results_list.append((
+                    0,
+                    model_name,
+                        {"tokens_per_second":0,"avg_time": 0, "input_tokens_count":0, "generated_tokens_count":0, "min_time": 0, "max_time": 0, "first_token_time": 0, "input_tokens_cost": 0.0, "generated_tokens_cost": 0.0, "status": f"✗ ({str(e)})"}
+                ))
+
+    # 按速度排序
+    results_list.sort(key=lambda x: x[0], reverse=True)
+    
+    # 添加排序后的结果到表格
+    for tokens_per_second, model_name, results in results_list:        
+        table.add_row(
+            model_name,  
+            f"{tokens_per_second:.2f}",
+            f"{results['first_token_time']:.2f}",
+            f"{results['input_tokens_count']}",
+            f"{results['generated_tokens_count']}",
+            f"{results['input_tokens_cost']:.4f}",
+            f"{results['generated_tokens_cost']:.4f}",            
+            results['status']
+        )        
+    
+    console.print(Panel(table, border_style="blue"))