Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!
import osos.environ["GEMINI_API_KEY"] = "your_gemini_api_key"from google import genaiclient = genai.Client()Available Models¶
sorted(
m.name for m in client.models.list()
)['models/aqa',
'models/deep-research-pro-preview-12-2025',
'models/gemini-2.0-flash',
'models/gemini-2.0-flash-001',
'models/gemini-2.0-flash-lite',
'models/gemini-2.0-flash-lite-001',
'models/gemini-2.5-computer-use-preview-10-2025',
'models/gemini-2.5-flash',
'models/gemini-2.5-flash-image',
'models/gemini-2.5-flash-lite',
'models/gemini-2.5-flash-native-audio-latest',
'models/gemini-2.5-flash-native-audio-preview-09-2025',
'models/gemini-2.5-flash-native-audio-preview-12-2025',
'models/gemini-2.5-flash-preview-tts',
'models/gemini-2.5-pro',
'models/gemini-2.5-pro-preview-tts',
'models/gemini-3-flash-preview',
'models/gemini-3-pro-image-preview',
'models/gemini-3-pro-preview',
'models/gemini-3.1-flash-image-preview',
'models/gemini-3.1-flash-lite-preview',
'models/gemini-3.1-flash-live-preview',
'models/gemini-3.1-flash-tts-preview',
'models/gemini-3.1-pro-preview',
'models/gemini-3.1-pro-preview-customtools',
'models/gemini-embedding-001',
'models/gemini-embedding-2-preview',
'models/gemini-flash-latest',
'models/gemini-flash-lite-latest',
'models/gemini-pro-latest',
'models/gemini-robotics-er-1.5-preview',
'models/gemini-robotics-er-1.6-preview',
'models/gemma-3-12b-it',
'models/gemma-3-1b-it',
'models/gemma-3-27b-it',
'models/gemma-3-4b-it',
'models/gemma-3n-e2b-it',
'models/gemma-3n-e4b-it',
'models/gemma-4-26b-a4b-it',
'models/gemma-4-31b-it',
'models/imagen-4.0-fast-generate-001',
'models/imagen-4.0-generate-001',
'models/imagen-4.0-ultra-generate-001',
'models/lyria-3-clip-preview',
'models/lyria-3-pro-preview',
'models/nano-banana-pro-preview',
'models/veo-2.0-generate-001',
'models/veo-3.0-fast-generate-001',
'models/veo-3.0-generate-001',
'models/veo-3.1-fast-generate-preview',
'models/veo-3.1-generate-preview',
'models/veo-3.1-lite-generate-preview']Prompts with File Context¶
file_md = client.files.upload(file="data/index.md")prompt = """
There are tables defined in HTML format in markdown files.
Help me convert each HTML table into a list-table in MyST Markdown.
Please refer to the official doc on list-table
at the link https://myst-parser.readthedocs.io/en/latest/syntax/tables.html#list-tables if necessary.
Below are some guidelines on the conversion.
- For both levels of the generated list-table,
I'd like to use `- `
which means that the very first row of the table starts with `- - `.
- Properly convert HTML hyperlinks defined using the `<a>` tag into markdown syntax.
For example,
`<a href="https://play.vercel.ai">AI Playground - Collection of LLMs</a>`
should be convert to
`[AI Playground - Collection of LLMs](https://play.vercel.ai)`
.
- Properly convert HTML superscript into footnote syntax in markdown.
For example,
`<sup>[1]</sup>`
should be convert to `[^1]`.
- Preserve `<span>` tags which define colored text.
A markdown file containing HTML tables to be processed is attached.
"""config = genai.types.GenerateContentConfig(
temperature=0,
system_instruction=(
"You are a file processing engine. "
"Output ONLY the final result of the processing. "
"Do not include any conversational text, introductory remarks, or concluding summaries."
)
)response = client.models.generate_content(
model="gemini-3.1-pro-preview",
contents=[prompt, file_md],
config=config,
)
print(response.text)---
title: Tips on Large Language Models
created: 2023-05-06 13:28:33
date: 2026-04-13 23:14:31.301515
authors:
- bendu
label: tips-on-large-language-models
license: CC-BY-4.0
tags:
- computer science
- programming
- AI
- machine learning
- LLM
- large language model
- Bard
- ChatGPT
---
**Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!**
```{list-table}
:header-rows: 1
- - Name
- Category
- API
- Description
- - [AI Playground - Collection of LLMs](https://play.vercel.ai)
-
- NA
-
- - [therundown.ai](https://therundown.ai)
- Education & Resource
-
- A company that provides AI education and resources, including a popular daily newsletter and an AI University. They aim to help individuals and businesses understand and utilize AI in their work and daily lives.
- - [consensus.app](https://consensus.app/)
- Research
-
- AI Search Engine for Research
- - [AI2 ScholarQA](https://scholarqa.allen.ai/chat)
- Research
-
-
- - [Emergent Mind](https://www.emergentmind.com/)
- Research
-
- AI Research Assistant for Computer Scientists
- - [DeepWiki](https://deepwiki.org/)
- Coding
- NA
-
- - [Google NotebookLM](https://notebooklm.google/)
- Knowledge Base
- NA
-
- - Tencent Yuanbao (in WeChat)
- Knowledge Base
- NA
- - can be added as a friend in WeChat
- - [otter.ai](https://otter.ai)
- Meeting
-
- An artificial intelligence (AI) meeting assistant that uses voice-to-text transcription to generate meeting notes, summaries, and transcripts. It can record and transcribe audio from various sources, including live conversations, recordings, and uploaded files, and offers features like speaker identification, inline image capture, and integration with popular platforms like Zoom and Google Meet.
- - [Midjourney](https://www.midjourney.com/explore?tab=top_month)
- Photo
-
-
- - [Google AI Studio](https://aistudio.google.com/)
- General
- Available
-
- - [Google Gemini CLI](https://www.legendu.net/misc/blog/tips-on-google-gemini-cli)
- General
- Available
- A command-line interface for Google Gemini.
- - [ChatGPT](https://chat.openai.com/)
- General
- Available
-
- - [happenstance.ai](https://happenstance.ai/)
- Social
-
- happenstance.ai is an AI-powered platform that helps users find people within their social networks. It uses AI to analyze your connected accounts (like LinkedIn, Gmail, and Twitter) and helps you discover hidden connections and individuals who match specific criteria. Essentially, it acts as a "ChatGPT for people," allowing you to search for individuals based on your network and desired qualifications.
- - [ollama](https://github.com/jmorganca/ollama)
- Deploy
-
- Get up and running with Llama 3.3, DeepSeek-R1, Phi-4, Gemma 3, Mistral Small 3.1 and other large language models.
- - [ToolBench](https://github.com/OpenBMB/ToolBench)
- Benchmark
-
- An open platform for training, serving, and evaluating large language model for tool learning.
```
[langextract](https://github.com/google/langextract)
## AI Tools from China
- [智谱AI开放平台](https://bigmodel.cn/)
- 腾讯元宝
- 豆包
## Tutorials
- [Stanford CS25: V2 I Introduction to Transformers w/ Andrej Karpathy](https://www.youtube.com/watch?v=XfpMkf4rD6E)
## Papers
- [Ai2 Scholar QA: Organized Literature Synthesis with Attribution](https://arxiv.org/abs/2504.10861)
## References
- [Tasks That AI Can Do for You](tasks-that-ai-can-do-for-you)
- [Tips on Google Gemini CLI](tips-on-google-gemini-cli)
- [Prompt Engineering for LLM Tools](prompt-engineering-for-llm-tools)
- [LLM in Rust](https://www.legendu.net/misc/blog/llm-in-rust)
- [deepwiki-open](https://github.com/AsyncFuncAI/deepwiki-open)