Show HN: I built an MCP that gives DeepSeek V4 Flash eyes (bad ones)

theJian1 pts0 comments

GitHub - theJian/squint-mcp: Most LLMs see images. With squint-mcp, the rest imagine seeing them. · GitHub

/" data-turbo-transient="true" />

Skip to content

Type / to search

Sign in<br>Sign upAppearance settings

You signed in with another tab or window. Reload to refresh your session.<br>You signed out in another tab or window. Reload to refresh your session.<br>You switched accounts on another tab or window. Reload to refresh your session.

Dismiss alert

{{ message }}

theJian

squint-mcp

Public

Notifications<br>You must be signed in to change notification settings

Fork

Star

main

BranchesTags

Go to file

CodeOpen more actions menu

Folders and files<br>NameNameLast commit message<br>Last commit date<br>Latest commit

History<br>53 Commits<br>53 Commits

eval

eval

run

run

src

src

.gitignore

.gitignore

.mcp.json

.mcp.json

CLAUDE.md

CLAUDE.md

README.md

README.md

bun.lock

bun.lock

opencode.json

opencode.json

package.json

package.json

tsconfig.json

tsconfig.json

View all files

Repository files navigation

(≖_≖ ) squint-mcp

Most LLMs see images. With squint-mcp, the rest imagine seeing them.

squint-mcp turns images into ASCII art for text-only LLMs (DeepSeek-V4-Flash,<br>GLM-5.2). They can then squint, crop, squint harder, and eventually announce that<br>your toaster is probably a building.

To use it:

"mcp": {<br>"squint": {<br>"type": "local",<br>"command": ["npx", "-y", "squint-mcp"]

Q & A

Is this vision?

Not really. It is ASCII art doing a vision impression.

Can it read tiny text?

It can try. Sometimes the text comes back as decorative soup.

Why not use a vision model?

That would solve the problem.

How is it different from a vision model?

It burns more tokens for worse results.

Is it any good?

No. But every now and then it gets something right,<br>which is much worse for everyone.

Contributing

If you make it less wonky, more useful, or better at distinguishing dogs from<br>furniture, pull requests are welcome.

License

Not decided yet. The code is young. Please be kind to it.

About<br>Most LLMs see images. With squint-mcp, the rest imagine seeing them.<br>Topics<br>ascii-graphicsdeepseek-v4-flashmcp-servertext-only-llm<br>Resources<br>Readme<br>Activity<br>Stars<br>0 stars<br>Watchers<br>0 watching<br>Forks<br>0 forks<br>Report repository

Releases

Packages

Contributors

Languages

You can’t perform that action at this time.

squint json llms images vision rest

Related Articles