{ "@context": "https://schema.org", "@id": "https://aggr.aymericbeaumet.com/items/kimi-com/2025-09-05-kimi-k2-instruct-0905/#webpage", "@type": [ "WebPage", "ArchiveComponent" ], "articleBody": "![Kimi K2: Open Agentic Intellignece](https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905/resolve/main/figures/kimi-logo.png)\n\n* * *\n\n[![Chat](https://img.shields.io/badge/🤖%20Chat-Kimi%20K2-ff6b6b?color=1783ff&logoColor=white)](https://www.kimi.com) [![github](https://img.shields.io/badge/🤖%20Github-Kimi%20K2-ff6b6b?color=1783ff&logoColor=white)](https://github.com/moonshotai/Kimi-K2) [![Homepage](https://img.shields.io/badge/Homepage-Moonshot%20AI-white?logo=Kimi&logoColor=white)](https://www.moonshot.ai)\n\n[![Hugging Face](https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Moonshot%20AI-ffc107?color=ffc107&logoColor=white)](https://huggingface.co/moonshotai) [![Twitter Follow](https://img.shields.io/badge/Twitter-Kimi.ai-white?logo=x&logoColor=white)](https://twitter.com/kimi_moonshot) [![Discord](https://img.shields.io/badge/Discord-Kimi.ai-white?logo=discord&logoColor=white)](https://discord.gg/TYU2fdJykW)\n\n[![License](https://img.shields.io/badge/License-Modified_MIT-f5de53?&color=f5de53)](https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905/blob/main/LICENSE)\n\n**📰  [Tech Blog](https://moonshotai.github.io/Kimi-K2/)**     |     **📄  [Paper](https://github.com/MoonshotAI/Kimi-K2/blob/main/tech_report.pdf)**\n\n## 1. Model Introduction\n\nKimi K2-Instruct-0905 is the latest, most capable version of Kimi K2. It is a state-of-the-art mixture-of-experts (MoE) language model, featuring 32 billion activated parameters and a total of 1 trillion parameters.\n\n### Key Features\n\n- Enhanced agentic coding intelligence: Kimi K2-Instruct-0905 demonstrates significant improvements in performance on public benchmarks and real-world coding agent tasks.\n- Improved frontend coding experience: Kimi K2-Instruct-0905 offers advancements in both the aesthetics and practicality of frontend programming.\n- Extended context length: Kimi K2-Instruct-0905’s context window has been increased from 128k to 256k tokens, providing better support for long-horizon tasks.\n\n## 2. Model Summary\n\n| | |\n| ------------------------------------------- | ------------------------ |\n| **Architecture** | Mixture-of-Experts (MoE) |\n| **Total Parameters** | 1T |\n| **Activated Parameters** | 32B |\n| **Number of Layers** (Dense layer included) | 61 |\n| **Number of Dense Layers** | 1 |\n| **Attention Hidden Dimension** | 7168 |\n| **MoE Hidden Dimension** (per Expert) | 2048 |\n| **Number of Attention Heads** | 64 |\n| **Number of Experts** | 384 |\n| **Selected Experts per Token** | 8 |\n| **Number of Shared Experts** | 1 |\n| **Vocabulary Size** | 160K |\n| **Context Length** | 256K |\n| **Attention Mechanism** | MLA |\n| **Activation Function** | SwiGLU |\n\n## 3. Evaluation Results\n\n| Benchmark | Metric | K2-Instruct-0905 | K2-Instruct-0711 | Qwen3-Coder-480B-A35B-Instruct | GLM-4.5 | DeepSeek-V3.1 | Claude-Sonnet-4 | Claude-Opus-4 |\n| ---------------------- | ------ | ---------------- | ---------------- | ------------------------------ | ------- | ------------- | --------------- | ------------- |\n| SWE-Bench verified | ACC | 69.2 ± 0.63 | 65.8 | 69.6\\* | 64.2\\* | 66.0\\* | 72.7\\* | 72.5\\* |\n| SWE-Bench Multilingual | ACC | 55.9 ± 0.72 | 47.3 | 54.7\\* | 52.7 | 54.5\\* | 53.3\\* | - |\n| Multi-SWE-Bench | ACC | 33.5 ± 0.28 | 31.3 | 32.7 | 31.7 | 29.0 | 35.7 | - |\n| Terminal-Bench | ACC | 44.5 ± 2.03 | 37.5 | 37.5\\* | 39.9\\* | 31.3\\* | 36.4\\* | 43.2\\* |\n| SWE-Dev | ACC | 66.6 ± 0.72 | 61.9 | 64.7 | 63.2 | 53.3 | 67.1 | - |\n\nAll K2-Instruct-0905 numbers are reported as mean ± std over five independent, full-test-set runs. Before each run we prune the repository so that every Git object unreachable from the target commit disappears; this guarantees the agent sees only the code that would legitimately be available at that point in history.\n\nExcept for Terminal-Bench (Terminus-2), every result was produced with our in-house evaluation harness. The harness is derived from SWE-agent, but we clamp the context windows of the Bash and Edit tools and rewrite the system prompt to match the task semantics. All baseline figures denoted with an asterisk (\\*) are excerpted directly from their official report or public leaderboard; the remaining metrics were evaluated by us under conditions identical to those used for K2-Instruct-0905.\n\nFor SWE-Dev we go one step further: we overwrite the original repository files and delete any test file that exercises the functions the agent is expected to generate, eliminating any indirect hints about the desired implementation.\n\n## 4. Deployment\n\n> You can access Kimi K2's API on [https://platform.moonshot.ai](https://platform.moonshot.ai) , we provide OpenAI/Anthropic-compatible API for you.\n>\n> The Anthropic-compatible API maps temperature by `real_temperature = request_temperature * 0.6` for better compatible with existing applications.\n\nOur model checkpoints are stored in the block-fp8 format, you can find it on [Huggingface](https://huggingface.co/moonshotai/Kimi-K2-Instruct).\n\nCurrently, Kimi-K2 is recommended to run on the following inference engines:\n\n- vLLM\n- SGLang\n- KTransformers\n- TensorRT-LLM\n\nDeployment examples for vLLM and SGLang can be found in the [Model Deployment Guide](https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905/blob/main/docs/deploy_guidance.md).\n\n* * *\n\n## 5. Model Usage\n\n### Chat Completion\n\nOnce the local inference service is up, you can interact with it through the chat endpoint:\n\n```python\ndef simple_chat(client: OpenAI, model_name: str):\n messages = [\n {\"role\": \"system\", \"content\": \"You are Kimi, an AI assistant created by Moonshot AI.\"},\n {\"role\": \"user\", \"content\": [{\"type\": \"text\", \"text\": \"Please give a brief self-introduction.\"}]},\n ]\n response = client.chat.completions.create(\n model=model_name,\n messages=messages,\n stream=False,\n temperature=0.6,\n max_tokens=256\n )\n print(response.choices[0].message.content)\n```\n\n> The recommended temperature for Kimi-K2-Instruct-0905 is `temperature = 0.6`. If no special instructions are required, the system prompt above is a good default.\n\n* * *\n\n### Tool Calling\n\nKimi-K2-Instruct-0905 has strong tool-calling capabilities. To enable them, you need to pass the list of available tools in each request, then the model will autonomously decide when and how to invoke them.\n\nThe following example demonstrates calling a weather tool end-to-end:\n\n```python\n# Your tool implementation\ndef get_weather(city: str) -> dict:\n return {\"weather\": \"Sunny\"}\n# Tool schema definition\ntools = [{\n \"type\": \"function\",\n \"function\": {\n \"name\": \"get_weather\",\n \"description\": \"Retrieve current weather information. Call this when the user asks about the weather.\",\n \"parameters\": {\n \"type\": \"object\",\n \"required\": [\"city\"],\n \"properties\": {\n \"city\": {\n \"type\": \"string\",\n \"description\": \"Name of the city\"\n }\n }\n }\n }\n}]\n# Map tool names to their implementations\ntool_map = {\n \"get_weather\": get_weather\n}\ndef tool_call_with_client(client: OpenAI, model_name: str):\n messages = [\n {\"role\": \"system\", \"content\": \"You are Kimi, an AI assistant created by Moonshot AI.\"},\n {\"role\": \"user\", \"content\": \"What's the weather like in Beijing today? Use the tool to check.\"}\n ]\n finish_reason = None\n while finish_reason is None or finish_reason == \"tool_calls\":\n completion = client.chat.completions.create(\n model=model_name,\n messages=messages,\n temperature=0.6,\n tools=tools, # tool list defined above\n tool_choice=\"auto\"\n )\n choice = completion.choices[0]\n finish_reason = choice.finish_reason\n if finish_reason == \"tool_calls\":\n messages.append(choice.message)\n for tool_call in choice.message.tool_calls:\n tool_call_name = tool_call.function.name\n tool_call_arguments = json.loads(tool_call.function.arguments)\n tool_function = tool_map[tool_call_name]\n tool_result = tool_function(**tool_call_arguments)\n print(\"tool_result:\", tool_result)\n messages.append({\n \"role\": \"tool\",\n \"tool_call_id\": tool_call.id,\n \"name\": tool_call_name,\n \"content\": json.dumps(tool_result)\n })\n print(\"-\" * 100)\n print(choice.message.content)\n```\n\nThe `tool_call_with_client` function implements the pipeline from user query to tool execution. This pipeline requires the inference engine to support Kimi-K2’s native tool-parsing logic. For more information, see the [Tool Calling Guide](https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905/blob/main/docs/tool_call_guidance.md).\n\n* * *\n\n## 6. License\n\nBoth the code repository and the model weights are released under the [Modified MIT License](https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905/tree/main/LICENSE).\n\n* * *\n\n## 7. Third Party Notices\n\nSee [THIRD PARTY NOTICES](https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905/blob/main/THIRD_PARTY_NOTICES.md)\n\n* * *\n\n## 7. Contact Us\n\nIf you have any questions, please reach out at [support@moonshot.cn](mailto:support@moonshot.cn).\n", "authors": [], "captured_at": "2026-10-19T01:26:53.998049939+00:00", "category": "ai", "content_html": "

\"Kimi

\n
\n

\"Chat\" \"github\" \"Homepage\"

\n

\"Hugging \"Twitter \"Discord\"

\n

\"License\"

\n

📰  Tech Blog     |     📄  Paper

\n

1. Model Introduction

\n

Kimi K2-Instruct-0905 is the latest, most capable version of Kimi K2. It is a state-of-the-art mixture-of-experts (MoE) language model, featuring 32 billion activated parameters and a total of 1 trillion parameters.

\n

Key Features

\n\n

2. Model Summary

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
ArchitectureMixture-of-Experts (MoE)
Total Parameters1T
Activated Parameters32B
Number of Layers (Dense layer included)61
Number of Dense Layers1
Attention Hidden Dimension7168
MoE Hidden Dimension (per Expert)2048
Number of Attention Heads64
Number of Experts384
Selected Experts per Token8
Number of Shared Experts1
Vocabulary Size160K
Context Length256K
Attention MechanismMLA
Activation FunctionSwiGLU
\n

3. Evaluation Results

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
BenchmarkMetricK2-Instruct-0905K2-Instruct-0711Qwen3-Coder-480B-A35B-InstructGLM-4.5DeepSeek-V3.1Claude-Sonnet-4Claude-Opus-4
SWE-Bench verifiedACC69.2 ± 0.6365.869.6*64.2*66.0*72.7*72.5*
SWE-Bench MultilingualACC55.9 ± 0.7247.354.7*52.754.5*53.3*-
Multi-SWE-BenchACC33.5 ± 0.2831.332.731.729.035.7-
Terminal-BenchACC44.5 ± 2.0337.537.5*39.9*31.3*36.4*43.2*
SWE-DevACC66.6 ± 0.7261.964.763.253.367.1-
\n

All K2-Instruct-0905 numbers are reported as mean ± std over five independent, full-test-set runs. Before each run we prune the repository so that every Git object unreachable from the target commit disappears; this guarantees the agent sees only the code that would legitimately be available at that point in history.

\n

Except for Terminal-Bench (Terminus-2), every result was produced with our in-house evaluation harness. The harness is derived from SWE-agent, but we clamp the context windows of the Bash and Edit tools and rewrite the system prompt to match the task semantics. All baseline figures denoted with an asterisk (*) are excerpted directly from their official report or public leaderboard; the remaining metrics were evaluated by us under conditions identical to those used for K2-Instruct-0905.

\n

For SWE-Dev we go one step further: we overwrite the original repository files and delete any test file that exercises the functions the agent is expected to generate, eliminating any indirect hints about the desired implementation.

\n

4. Deployment

\n
\n

You can access Kimi K2's API on https://platform.moonshot.ai , we provide OpenAI/Anthropic-compatible API for you.

\n

The Anthropic-compatible API maps temperature by real_temperature = request_temperature * 0.6 for better compatible with existing applications.

\n
\n

Our model checkpoints are stored in the block-fp8 format, you can find it on Huggingface.

\n

Currently, Kimi-K2 is recommended to run on the following inference engines:

\n\n

Deployment examples for vLLM and SGLang can be found in the Model Deployment Guide.

\n
\n

5. Model Usage

\n

Chat Completion

\n

Once the local inference service is up, you can interact with it through the chat endpoint:

\n
def simple_chat(client: OpenAI, model_name: str):\n    messages = [\n        {"role": "system", "content": "You are Kimi, an AI assistant created by Moonshot AI."},\n        {"role": "user", "content": [{"type": "text", "text": "Please give a brief self-introduction."}]},\n    ]\n    response = client.chat.completions.create(\n        model=model_name,\n        messages=messages,\n        stream=False,\n        temperature=0.6,\n        max_tokens=256\n    )\n    print(response.choices[0].message.content)\n
\n
\n

The recommended temperature for Kimi-K2-Instruct-0905 is temperature = 0.6. If no special instructions are required, the system prompt above is a good default.

\n
\n
\n

Tool Calling

\n

Kimi-K2-Instruct-0905 has strong tool-calling capabilities. To enable them, you need to pass the list of available tools in each request, then the model will autonomously decide when and how to invoke them.

\n

The following example demonstrates calling a weather tool end-to-end:

\n
# Your tool implementation\ndef get_weather(city: str) -> dict:\n    return {"weather": "Sunny"}\n# Tool schema definition\ntools = [{\n    "type": "function",\n    "function": {\n        "name": "get_weather",\n        "description": "Retrieve current weather information. Call this when the user asks about the weather.",\n        "parameters": {\n            "type": "object",\n            "required": ["city"],\n            "properties": {\n                "city": {\n                    "type": "string",\n                    "description": "Name of the city"\n                }\n            }\n        }\n    }\n}]\n# Map tool names to their implementations\ntool_map = {\n    "get_weather": get_weather\n}\ndef tool_call_with_client(client: OpenAI, model_name: str):\n    messages = [\n        {"role": "system", "content": "You are Kimi, an AI assistant created by Moonshot AI."},\n        {"role": "user", "content": "What's the weather like in Beijing today? Use the tool to check."}\n    ]\n    finish_reason = None\n    while finish_reason is None or finish_reason == "tool_calls":\n        completion = client.chat.completions.create(\n            model=model_name,\n            messages=messages,\n            temperature=0.6,\n            tools=tools,          # tool list defined above\n            tool_choice="auto"\n        )\n        choice = completion.choices[0]\n        finish_reason = choice.finish_reason\n        if finish_reason == "tool_calls":\n            messages.append(choice.message)\n            for tool_call in choice.message.tool_calls:\n                tool_call_name = tool_call.function.name\n                tool_call_arguments = json.loads(tool_call.function.arguments)\n                tool_function = tool_map[tool_call_name]\n                tool_result = tool_function(**tool_call_arguments)\n                print("tool_result:", tool_result)\n                messages.append({\n                    "role": "tool",\n                    "tool_call_id": tool_call.id,\n                    "name": tool_call_name,\n                    "content": json.dumps(tool_result)\n                })\n    print("-" * 100)\n    print(choice.message.content)\n
\n

The tool_call_with_client function implements the pipeline from user query to tool execution. This pipeline requires the inference engine to support Kimi-K2’s native tool-parsing logic. For more information, see the Tool Calling Guide.

\n
\n

6. License

\n

Both the code repository and the model weights are released under the Modified MIT License.

\n
\n

7. Third Party Notices

\n

See THIRD PARTY NOTICES

\n
\n

7. Contact Us

\n

If you have any questions, please reach out at support@moonshot.cn.

\n", "content_markdown": "![Kimi K2: Open Agentic Intellignece](https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905/resolve/main/figures/kimi-logo.png)\n\n* * *\n\n[![Chat](https://img.shields.io/badge/🤖%20Chat-Kimi%20K2-ff6b6b?color=1783ff&logoColor=white)](https://www.kimi.com) [![github](https://img.shields.io/badge/🤖%20Github-Kimi%20K2-ff6b6b?color=1783ff&logoColor=white)](https://github.com/moonshotai/Kimi-K2) [![Homepage](https://img.shields.io/badge/Homepage-Moonshot%20AI-white?logo=Kimi&logoColor=white)](https://www.moonshot.ai)\n\n[![Hugging Face](https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Moonshot%20AI-ffc107?color=ffc107&logoColor=white)](https://huggingface.co/moonshotai) [![Twitter Follow](https://img.shields.io/badge/Twitter-Kimi.ai-white?logo=x&logoColor=white)](https://twitter.com/kimi_moonshot) [![Discord](https://img.shields.io/badge/Discord-Kimi.ai-white?logo=discord&logoColor=white)](https://discord.gg/TYU2fdJykW)\n\n[![License](https://img.shields.io/badge/License-Modified_MIT-f5de53?&color=f5de53)](https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905/blob/main/LICENSE)\n\n**📰  [Tech Blog](https://moonshotai.github.io/Kimi-K2/)**     |     **📄  [Paper](https://github.com/MoonshotAI/Kimi-K2/blob/main/tech_report.pdf)**\n\n## 1. Model Introduction\n\nKimi K2-Instruct-0905 is the latest, most capable version of Kimi K2. It is a state-of-the-art mixture-of-experts (MoE) language model, featuring 32 billion activated parameters and a total of 1 trillion parameters.\n\n### Key Features\n\n- Enhanced agentic coding intelligence: Kimi K2-Instruct-0905 demonstrates significant improvements in performance on public benchmarks and real-world coding agent tasks.\n- Improved frontend coding experience: Kimi K2-Instruct-0905 offers advancements in both the aesthetics and practicality of frontend programming.\n- Extended context length: Kimi K2-Instruct-0905’s context window has been increased from 128k to 256k tokens, providing better support for long-horizon tasks.\n\n## 2. Model Summary\n\n| | |\n| ------------------------------------------- | ------------------------ |\n| **Architecture** | Mixture-of-Experts (MoE) |\n| **Total Parameters** | 1T |\n| **Activated Parameters** | 32B |\n| **Number of Layers** (Dense layer included) | 61 |\n| **Number of Dense Layers** | 1 |\n| **Attention Hidden Dimension** | 7168 |\n| **MoE Hidden Dimension** (per Expert) | 2048 |\n| **Number of Attention Heads** | 64 |\n| **Number of Experts** | 384 |\n| **Selected Experts per Token** | 8 |\n| **Number of Shared Experts** | 1 |\n| **Vocabulary Size** | 160K |\n| **Context Length** | 256K |\n| **Attention Mechanism** | MLA |\n| **Activation Function** | SwiGLU |\n\n## 3. Evaluation Results\n\n| Benchmark | Metric | K2-Instruct-0905 | K2-Instruct-0711 | Qwen3-Coder-480B-A35B-Instruct | GLM-4.5 | DeepSeek-V3.1 | Claude-Sonnet-4 | Claude-Opus-4 |\n| ---------------------- | ------ | ---------------- | ---------------- | ------------------------------ | ------- | ------------- | --------------- | ------------- |\n| SWE-Bench verified | ACC | 69.2 ± 0.63 | 65.8 | 69.6\\* | 64.2\\* | 66.0\\* | 72.7\\* | 72.5\\* |\n| SWE-Bench Multilingual | ACC | 55.9 ± 0.72 | 47.3 | 54.7\\* | 52.7 | 54.5\\* | 53.3\\* | - |\n| Multi-SWE-Bench | ACC | 33.5 ± 0.28 | 31.3 | 32.7 | 31.7 | 29.0 | 35.7 | - |\n| Terminal-Bench | ACC | 44.5 ± 2.03 | 37.5 | 37.5\\* | 39.9\\* | 31.3\\* | 36.4\\* | 43.2\\* |\n| SWE-Dev | ACC | 66.6 ± 0.72 | 61.9 | 64.7 | 63.2 | 53.3 | 67.1 | - |\n\nAll K2-Instruct-0905 numbers are reported as mean ± std over five independent, full-test-set runs. Before each run we prune the repository so that every Git object unreachable from the target commit disappears; this guarantees the agent sees only the code that would legitimately be available at that point in history.\n\nExcept for Terminal-Bench (Terminus-2), every result was produced with our in-house evaluation harness. The harness is derived from SWE-agent, but we clamp the context windows of the Bash and Edit tools and rewrite the system prompt to match the task semantics. All baseline figures denoted with an asterisk (\\*) are excerpted directly from their official report or public leaderboard; the remaining metrics were evaluated by us under conditions identical to those used for K2-Instruct-0905.\n\nFor SWE-Dev we go one step further: we overwrite the original repository files and delete any test file that exercises the functions the agent is expected to generate, eliminating any indirect hints about the desired implementation.\n\n## 4. Deployment\n\n> You can access Kimi K2's API on [https://platform.moonshot.ai](https://platform.moonshot.ai) , we provide OpenAI/Anthropic-compatible API for you.\n>\n> The Anthropic-compatible API maps temperature by `real_temperature = request_temperature * 0.6` for better compatible with existing applications.\n\nOur model checkpoints are stored in the block-fp8 format, you can find it on [Huggingface](https://huggingface.co/moonshotai/Kimi-K2-Instruct).\n\nCurrently, Kimi-K2 is recommended to run on the following inference engines:\n\n- vLLM\n- SGLang\n- KTransformers\n- TensorRT-LLM\n\nDeployment examples for vLLM and SGLang can be found in the [Model Deployment Guide](https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905/blob/main/docs/deploy_guidance.md).\n\n* * *\n\n## 5. Model Usage\n\n### Chat Completion\n\nOnce the local inference service is up, you can interact with it through the chat endpoint:\n\n```python\ndef simple_chat(client: OpenAI, model_name: str):\n messages = [\n {\"role\": \"system\", \"content\": \"You are Kimi, an AI assistant created by Moonshot AI.\"},\n {\"role\": \"user\", \"content\": [{\"type\": \"text\", \"text\": \"Please give a brief self-introduction.\"}]},\n ]\n response = client.chat.completions.create(\n model=model_name,\n messages=messages,\n stream=False,\n temperature=0.6,\n max_tokens=256\n )\n print(response.choices[0].message.content)\n```\n\n> The recommended temperature for Kimi-K2-Instruct-0905 is `temperature = 0.6`. If no special instructions are required, the system prompt above is a good default.\n\n* * *\n\n### Tool Calling\n\nKimi-K2-Instruct-0905 has strong tool-calling capabilities. To enable them, you need to pass the list of available tools in each request, then the model will autonomously decide when and how to invoke them.\n\nThe following example demonstrates calling a weather tool end-to-end:\n\n```python\n# Your tool implementation\ndef get_weather(city: str) -> dict:\n return {\"weather\": \"Sunny\"}\n# Tool schema definition\ntools = [{\n \"type\": \"function\",\n \"function\": {\n \"name\": \"get_weather\",\n \"description\": \"Retrieve current weather information. Call this when the user asks about the weather.\",\n \"parameters\": {\n \"type\": \"object\",\n \"required\": [\"city\"],\n \"properties\": {\n \"city\": {\n \"type\": \"string\",\n \"description\": \"Name of the city\"\n }\n }\n }\n }\n}]\n# Map tool names to their implementations\ntool_map = {\n \"get_weather\": get_weather\n}\ndef tool_call_with_client(client: OpenAI, model_name: str):\n messages = [\n {\"role\": \"system\", \"content\": \"You are Kimi, an AI assistant created by Moonshot AI.\"},\n {\"role\": \"user\", \"content\": \"What's the weather like in Beijing today? Use the tool to check.\"}\n ]\n finish_reason = None\n while finish_reason is None or finish_reason == \"tool_calls\":\n completion = client.chat.completions.create(\n model=model_name,\n messages=messages,\n temperature=0.6,\n tools=tools, # tool list defined above\n tool_choice=\"auto\"\n )\n choice = completion.choices[0]\n finish_reason = choice.finish_reason\n if finish_reason == \"tool_calls\":\n messages.append(choice.message)\n for tool_call in choice.message.tool_calls:\n tool_call_name = tool_call.function.name\n tool_call_arguments = json.loads(tool_call.function.arguments)\n tool_function = tool_map[tool_call_name]\n tool_result = tool_function(**tool_call_arguments)\n print(\"tool_result:\", tool_result)\n messages.append({\n \"role\": \"tool\",\n \"tool_call_id\": tool_call.id,\n \"name\": tool_call_name,\n \"content\": json.dumps(tool_result)\n })\n print(\"-\" * 100)\n print(choice.message.content)\n```\n\nThe `tool_call_with_client` function implements the pipeline from user query to tool execution. This pipeline requires the inference engine to support Kimi-K2’s native tool-parsing logic. For more information, see the [Tool Calling Guide](https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905/blob/main/docs/tool_call_guidance.md).\n\n* * *\n\n## 6. License\n\nBoth the code repository and the model weights are released under the [Modified MIT License](https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905/tree/main/LICENSE).\n\n* * *\n\n## 7. Third Party Notices\n\nSee [THIRD PARTY NOTICES](https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905/blob/main/THIRD_PARTY_NOTICES.md)\n\n* * *\n\n## 7. Contact Us\n\nIf you have any questions, please reach out at [support@moonshot.cn](mailto:support@moonshot.cn).\n", "dateCreated": "2026-10-19T01:26:53.998049939+00:00", "date_created": "2025-09-05T00:00:00+00:00", "date_modified": null, "date_published": "2025-09-05T00:00:00+00:00", "description": "2025-09-05", "external_url": "https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905", "git": { "commit": "ebb0941c5c06a2ab2377342d607535262a3e969e", "path": "items/kimi-com/2025/09/2025-09-05-kimi-k2-instruct-0905.md", "permalink": "https://github.com/aymericbeaumet/aggr-instance/blob/ebb0941c5c06a2ab2377342d607535262a3e969e/items/kimi-com/2025/09/2025-09-05-kimi-k2-instruct-0905.md" }, "id": "urn:aggr:item:6f86fe9c1cacde2c74464ac78b2ed523b48fe9ba", "isBasedOn": { "@id": "https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905" }, "mainEntity": { "@id": "https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905", "@type": "CreativeWork", "archivedAt": { "@id": "https://aggr.aymericbeaumet.com/items/kimi-com/2025-09-05-kimi-k2-instruct-0905/#webpage" }, "datePublished": "2025-09-05T00:00:00+00:00", "headline": "Kimi-K2-Instruct-0905", "image": { "@type": "ImageObject", "caption": "Kimi-K2-Instruct-0905", "height": 90, "url": "https://kimi-file.kimi.ai/prod-chat-kimi/kfs/4/2/2026-03-25/1d71tdsudcmosb3tpmud0?x-tos-process=image%2Fauto-orient%2C1%2Fstrip%2Fignore-error%2C1", "width": 160 }, "timeRequired": "PT4M", "url": "https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905", "wordCount": 888 }, "name": "Archived snapshot: Kimi-K2-Instruct-0905", "preview": { "alt": "Kimi-K2-Instruct-0905", "color": null, "height": 90, "placeholder": { "data_url": "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAQAAAC1HAwCAAAAC0lEQVR42mNk+A8AAQUBAScY42YAAAAASUVORK5CYII=", "hash": "" }, "url": "https://kimi-file.kimi.ai/prod-chat-kimi/kfs/4/2/2026-03-25/1d71tdsudcmosb3tpmud0?x-tos-process=image%2Fauto-orient%2C1%2Fstrip%2Fignore-error%2C1", "width": 160 }, "reading_minutes": 4, "replicated_at": null, "source": "kimi-com", "source_name": "Kimi Research Blog | Moonshot AI", "summary": "2025-09-05", "tags": [], "temporalCoverage": "2026-10-19T01:26:53.998049939+00:00", "title": "Kimi-K2-Instruct-0905", "url": "https://aggr.aymericbeaumet.com/items/kimi-com/2025-09-05-kimi-k2-instruct-0905/", "word_count": 888 }