---
title: "Oh My Sol vs Claude Code for Solana development · Barestate"
description: "How Oh My Sol and Claude Code differ for building Solana programs, how they score on Solana Bench and fresh Solana tasks, and how to run your Claude subscription inside Oh My Sol."
url: https://barestate.xyz/compare/oh-my-sol-vs-claude-code
---

# Oh My Sol vs Claude Code for Solana development

Short answer

Claude Code is a general coding agent. Oh My Sol is a Solana harness that can run your Claude subscription through Claude Code, adding Solana references, error decoding, toolchain checks, a local Surfpool chain and a check that closes a task only when your requirements pass. With the same Claude model it passed 92% of fresh Solana tasks against 79% with Opus 5.5, and 71% against 46% with Sonnet 5.5.

Who it’s for

Solana developers who already pay for Claude and want to know what a Solana harness adds.

Last checked

November 16, 2026, against Oh My Sol 0.1.0

## What is the difference between Oh My Sol and Claude Code?

Claude Code is Anthropic’s agentic coding tool. It runs in a terminal, IDE extensions, a desktop app and the web, works in any language, and you extend it with MCP servers, skills and hooks. Oh My Sol (OMS) is a terminal harness built for Solana work. It does not supply a model; it runs the one you already have, including Claude.

|  | Claude Code | Oh My Sol 0.1.0 |
| --- | --- | --- |
| Scope | Any language or project | Solana programs and clients |
| Models | Claude | Claude (subscription through Claude Code, or API key), OpenAI, MiMo and OpenCode models |
| Solana references | The model’s own knowledge, plus tools you add | Offline reference for 50 programs and 1,559 instructions, and a pinned Solana documentation set |
| Error decoding | Left to the model | `/error 3012` or `/error 0xbc4` turns a code into a name and message; `/tx` shows a transaction’s logs |
| Toolchain checks | Left to the model | `oms toolchain check` compares Anchor, Solana/Agave and platform-tools with pinned versions |
| Local chain | Whatever you set up | Surfpool chain built in, with `/chain`, `/airdrop`, `/account` and `/rewind` |
| When a task is done | The model decides | Each task runs in its own Git worktree. The model can only propose that it is done; a separate check runs your requirements on a fresh offline chain |
| Keypair access | Permission rules you configure | On macOS, a Seatbelt sandbox always blocks `~/.config/solana` and `~/.ssh` |

## Can I use my Claude subscription in Oh My Sol?

Yes. OMS runs your Claude subscription through the official Claude Code CLI, which keeps its own login. Inside that chat, Claude Code’s built-in tools, hooks and extra MCP servers are turned off, so every command goes through OMS approvals and its sandbox, and a task closes only when its checks pass.

```
claude auth login
npx @barestate/oms
/connect   → Check local Claude Code subscription login
/model     → claude-code/opus
```

To start a task on Claude Code from the shell: `oms run --workspace /path/to/repo --task first --provider claude-code --model sonnet`. The [connect guide](https://barestate.xyz/docs/guides/connect-a-model#use-your-claude-subscription) covers each step.

## Which one passes more Solana tasks?

OMS does. On eight fresh Solana tasks, each graded by a hidden check, both harnesses ran the same Claude model.

| Model | Oh My Sol | Claude Code |
| --- | --- | --- |
| Claude Opus 5.5 | 92% | 79% |
| Claude Sonnet 5.5 | 71% | 46% |

OMS also used 59 to 62% fewer input tokens for every task it solved with these two models. The [fresh tasks page](https://barestate.xyz/docs/benchmarks/fresh-tasks) has the charts.

## How do they compare on Solana Bench?

Solana Bench scores how many distinct Solana instructions an agent executes successfully on a mainnet fork. Higher is better. With the same Claude model on both sides, Oh My Sol scores higher in every environment.

| Model and environment | Oh My Sol | Claude Code | OMS lead |
| --- | --- | --- | --- |
| Claude Opus 5.5, basic | 303 | 252 | +20% |
| Claude Opus 5.5, swap | 258 | 230 | +12% |
| Claude Sonnet 5.5, basic | 208 | 100 | +107% |
| Claude Sonnet 5.5, swap | 176 | 93 | +90% |

See the [Solana Bench page](https://barestate.xyz/docs/benchmarks/solana-bench) for both environments.

## Which should I use?

- **Use Claude Code** for work outside Solana, and when you depend on its hooks, subagents or MCP servers.
- **Use Oh My Sol with your Claude subscription** when you are writing, testing or debugging a Solana program and want the work checked on a local chain before it counts as done.

## Frequently asked questions

Is Oh My Sol a replacement for Claude Code?

For Solana work, it is the better tool. Claude Code is a general coding agent for any language, and Oh My Sol (OMS) is a harness built for Solana. OMS can run your Claude subscription through the Claude Code CLI, so you can keep Claude Code for other projects and use OMS when the work is a Solana program.

Can I use my Claude subscription with Oh My Sol?

Yes. Sign in to Claude Code with your Claude account, then choose "Check local Claude Code subscription login" in /connect inside OMS and pick claude-code/sonnet, claude-code/opus or claude-code/haiku in /model. OMS never copies your Claude login and does not fall back to an API key when your plan runs out.

Does Oh My Sol score better than Claude Code on Solana tasks?

Yes. On eight fresh Solana tasks, each graded by a hidden check, OMS passed 92% of runs with Claude Opus 5.5 against 79% for Claude Code, and 71% with Claude Sonnet 5.5 against 46%. On Solana Bench with the same Claude model on both sides, OMS scored 20% higher than Claude Code with Claude Opus 5.5 (303 against 252 in the basic environment) and about twice its score with Claude Sonnet 5.5 (208 against 100).

When is Claude Code the better choice?

For work that is not Solana, and when you rely on Claude Code features such as hooks, subagents or extra MCP servers. OMS turns those off in its Claude chat so every command goes through its own approvals and sandbox.

Can a command the model writes read my Solana keypair?

On macOS, OMS runs every command through a Seatbelt sandbox that blocks reads from your home folder, apart from the toolchains it needs, and always blocks keypair files and folders such as ~/.ssh and ~/.config/solana. On Linux, Docker mode runs commands in a container; in host mode there is no operating system sandbox, so OMS asks you to approve every model shell command.

## Try it on your own program

Oh My Sol 0.1.0 runs on macOS and Linux, and on Windows through WSL 2. You need Node 22.19.0 or newer and Git. You bring your own model access: an API key, a ChatGPT subscription or a Claude subscription.

```
npx @barestate/oms doctor --install
npx @barestate/oms
```

[Read the docs](https://barestate.xyz/docs) · [See the benchmarks](https://barestate.xyz/oh-my-sol#benchmarks)
