Run aider on any BlockRun model through a local OpenAI-compatible sidecar
Point aider's OpenAI base URL at the blockrun-litellm sidecar and it pair-programs on any model in the catalog — switch models with one flag, no provider key, pay per call.
How it works
aider talks to models through LiteLLM and reads an OpenAI-compatible base URL from its environment. The blockrun-litellm sidecar is exactly that: a local proxy that signs an x402 payment for each request from your wallet, so aider needs no provider key and can name any catalog model with the openai/ prefix aider expects for a custom endpoint.
Switching the editing model is one flag. A cheap model for the routine edits and a frontier model for the architecture pass both come from the same endpoint, and aider's own repo map, diffs and git commits are unchanged — the sidecar is transparent to everything above it.
Behind it is the BlockRun catalog: 78 chat models plus image, video, music and speech generation, search, market data and multi-chain RPC, priced per call. 6 open-weight models are free and need no wallet.
Setup
1. Start the sidecar
pip install 'blockrun-litellm[proxy]'\nexport BLOCKRUN_WALLET_KEY=0x<base wallet key>\nblockrun-litellm-proxy --port 40012. Point aider at it
export OPENAI_API_BASE=http://127.0.0.1:4001/v1
export OPENAI_API_KEY=dummy
aider --model openai/anthropic/claude-sonnet-5Questions
- How do I use aider with BlockRun?
- Start the blockrun-litellm sidecar, export OPENAI_API_BASE to it, and pass any catalog model to aider with the openai/ prefix aider uses for custom endpoints, as in the setup above.
- Which model should I give aider?
- Any on the catalog. The model id after the openai/ prefix is the BlockRun id, so anthropic/claude-sonnet-5 and openai/gpt-5.4 are both valid choices for the same flag.
- Does aider's architect mode work with two BlockRun models?
- Yes. Set the architect and editor models to two catalog ids; both go through the same sidecar and are paid per call.
- Is the sidecar exposed to my network?
- It binds to loopback by default; set BLOCKRUN_PROXY_TOKEN before binding anywhere else.
- aider on BlockRun vs aider with a provider key?
- Same aider, same features. The difference is one wallet for every provider instead of a key per provider, and a per-call price you see before the call instead of a monthly bill after it.
- What does a call cost through this integration?
- The provider's published token rate with no markup on tokens, plus a flat per-call fee, quoted in dollars in the 402 before the call runs. Image, video and speech calls are priced per unit the same way. The live rate card is on the models page.
- Do I need an API key or an account?
- No. On the pay-per-call door the request carries its own payment from a wallet the SDK or MCP server creates on first run; there is nothing to sign up for. Teams that prefer keys get an API key and a monthly invoice instead.