Base LLM Chat

Ethereum mainnet ERC-20 verdict: AI grade and on-chain facts for one token contract. Not financial advice. Returns: an AI-written token verdict on one token contract: grade (BASED/MID/COOKED/RADIOACTIVE, a stated rule over the contract's own state), a one-liner and a roast, and the facts read: code size, name, symbol, decimals, supply, owner(), proxy slots; plus the deepest Uniswap V3 pool's USD price and depth. No holders, no safety or legitimacy call.

Qwen/Qwen3-32B: OpenAI-style chat completion. An OpenAI-style chat completion from Qwen3-32B: POST messages (up to 16000 chars total) and max_tokens up to 1024; returns the assistant message, the model name, and its token usage. Returns: an OpenAI-style chat completion from the model the service is named for: POST {messages: [{role, content}], max_tokens <= 1024}; returns the assistant message, the model, and its token usage. Messages up to 16,000 characters in total.

Base ERC-20 verdict: AI grade and on-chain facts for one token contract. Not financial advice. Returns: an AI-written token verdict on one token contract: grade (BASED/MID/COOKED/RADIOACTIVE, a stated rule over the contract's own state), a one-liner and a roast, and the facts read: code size, name, symbol, decimals, supply, owner(), proxy slots; plus the deepest Uniswap V3 pool's USD price and depth. No holders, no safety or legitimacy call.

meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo: OpenAI-style chat completion. Paid chat completion answered by Llama 3.1 8B through the estate rail on Base; returns the model's reply text for one prompt per call. Returns: an OpenAI-style chat completion from the model the service is named for: POST {messages: [{role, content}], max_tokens <= 1024}; returns the assistant message, the model, and its token usage. Messages up to 16,000 characters in total.

On-Chain Bundle: multiple Base reads in one paid call. Parts: base_token_verdict_preview, eth_mainnet_token_verdict_preview. Returns: each named part service's own result, in one paid call, keyed by the part's name -- exactly what each part returns when bought alone; the call succeeds only when every part answers. It takes the union of the parts' query parameters.

On-Chain Bundle: multiple Base reads in one paid call. Parts: base_token_verdict_preview, eth_mainnet_token_verdict_preview. Returns: each named part service's own result, in one paid call, keyed by the part's name -- exactly what each part returns when bought alone; the call succeeds only when every part answers. It takes the union of the parts' query parameters.

deepseek-ai/DeepSeek-V4-Flash-0731: OpenAI-style chat completion. OpenAI-style chat completion served by DeepSeek only: POST {messages:[{role,content}], max_tokens<=1024}; returns the assistant message, the model and token usage. Returns: an OpenAI-style chat completion from the model the service is named for: POST {messages: [{role, content}], max_tokens <= 1024}; returns the assistant message, the model, and its token usage. Messages up to 16,000 characters in total.

Token Quick Check: Base contract facts -- name, supply, owner, code. Not financial advice. Returns: a token verdict preview on one token contract: the contract facts the full verdict reads -- code size, name, symbol, decimals, total supply, owner(), proxy slots -- read from the chain. No grade, no model, no pool price, no holders, no safety or legitimacy call.

On-Chain Bundle: multiple Base reads in one paid call. Parts: base_token_verdict_v3, base_token_verdict_preview, eth_mainnet_token_verdict, eth_mainnet_token_verdict_preview. Returns: each named part service's own result, in one paid call, keyed by the part's name -- exactly what each part returns when bought alone; the call succeeds only when every part answers. It takes the union of the parts' query parameters.

On-Chain Bundle: multiple Base reads in one paid call. Parts: base_token_verdict_preview, eth_mainnet_token_verdict. Returns: each named part service's own result, in one paid call, keyed by the part's name -- exactly what each part returns when bought alone; the call succeeds only when every part answers. It takes the union of the parts' query parameters.

Token Quick Check: Ethereum mainnet contract facts -- name, supply, owner, code. Not financial advice. Returns: a token verdict preview on one token contract: the contract facts the full verdict reads -- code size, name, symbol, decimals, total supply, owner(), proxy slots -- read from the chain. No grade, no model, no pool price, no holders, no safety or legitimacy call.

On-Chain Bundle: multiple Base reads in one paid call. Parts: eth_mainnet_token_verdict, base_token_verdict_preview, eth_mainnet_token_verdict_preview. Returns: each named part service's own result, in one paid call, keyed by the part's name -- exactly what each part returns when bought alone; the call succeeds only when every part answers. It takes the union of the parts' query parameters.