New blocks

gemini-2.0-flash-thinking-exp-01-21

gemini 2.0 flash thinking exp 01-21

React Hook Form

Summarize market research

My Custom HTTP Server

shadcnblocks

Shadcn UI blocks

llama-3.1-8b-instruct

Meta's Llama 3.1 8B-instruct model combines efficient size with strong reasoning and multilingual capabilities. This instruction-tuned LLM delivers powerful language understanding and generation in a compact 8B parameter package.

Voyage Rerank 2

Voyage's generalist reranker optimized for quality with multilingual support

Voyage Code 3

32 Batch Size and 16k chunk size (forced by Rerank 2; feel free to Remix this and set to 32k in case you do not use Rerank 2).

DeepSeek R1

Used for non-autocomplete tasks and reasoning.

DeepSeek V3 Autocomplete

Set with temperature=0, ensuring context and generated tokens count to be low, ideal for autocomplete.

ReactFlow

React Hook Form

Vercel AI SDK

qwen2.5-coder 1.5b

The latest series of Code-Specific Qwen models, with significant improvements in code generation, code reasoning, and code fixing

Qwen 2.5 Coder 7B

The latest series of Code-Specific Qwen models, with significant improvements in code generation, code reasoning, and code fixing

Bun Guides

Bun list of examples

Bun

Bun Documentation

Vercel Docs

Chinese Lang Rules

说中文

Vercel Examples

The git repo of Vercel Examples

llama-3.1-8b-instruct

Meta's latest class of models, Llama 3.1, launched with a variety of sizes and configurations. The 8B instruct-tuned version is particularly fast and efficient. It has demonstrated strong performance in human evaluations, outperforming several leading closed-source models.

Vercel Edge Runtime

Deepseek-r1-distill-llama-70b

DeepSeek R1 Distill LLama 70B

Dotenv docs

The secrets manager for .env files – from the same people that pioneered dotenv.

codegemma

Gemma 2

bge-reranker-v2-m3

Claude 3.5 Sonnet

Anthropic's previous most intelligent model

deepseek-r1 8b

DeepSeek-R1 achieves performance comparable to OpenAI-o1 across math, code, and reasoning tasks

qwen2.5-coder 3b

deepseek-r1

DeepSeek-R1 achieves performance comparable to OpenAI-o1 across math, code, and reasoning tasks

deepseek-r1 8b

DeepSeek-R1 achieves performance comparable to OpenAI-o1 across math, code, and reasoning tasks

Foundry Documentation

Foundry is a blazing fast, portable and modular toolkit for Ethereum application development written in Rust. Foundry consists of: - Forge: Build, test, fuzz, debug and deploy Solidity contracts, like Hardhat, Brownie, Ape. - Cast: A Swiss Army knife for interacting with EVM smart contracts, sending transactions and getting chain data. - Anvil: Fast local Ethereum development node, akin to Hardhat Network, Tenderly. - Chisel: Fast, utilitarian, and verbose Solidity REPL.

Use Unified Diff for Apply

This block can be used to encourage models to use a Unified Diff format when generating code blocks off of existing code. Using the Apply feature with this format will be quicker and more token efficient.

Jupyter learning rules

PyTorch Rules