Ben Klieger
@benklieger
Followers
4K
Following
3K
Media
205
Statuses
1K
compound ai lead @groqinc, cs & research @stanford | opinions are my own
Newest projects on Github
Joined January 2023
Connect Kimi K2 to live docs for 30,000 popular frameworks 🔥. Most AI coding assistants, while impressive, have a significant gap — outdated implementation knowledge. I spend this weekend building a new open source project to fix that: Meet CodeWizard, Kimi K2 running fast on
7
23
134
Exciting to see customers succeed with compound on @GroqInc!.
Congrats to @getRecallAI for hitting #1 on Product Hunt this week powered in part by @GroqInc and Compound Beta
1
1
21
More evals!.
OpenBench v0.2.0 is here 🚀. Big coverage jump: 17 new benchmarks across math, reasoning, reading comp, health, long-context recall, plus first-class support for local evals. We also have full OpenAI Simple-Evals parity!
0
0
7
These apply to Gemini code, qwen code, groq code, etc… genuinely would be huge quality of life improvements. Cc @alexalbert__ @lee_x64.
0
0
3
Claude code feature requests:. 1. Notification sound whenever Claude is asking for permission to do something, so I know to go back to the window. 2. A mode where Claude code can make changes across files and let approval happen at the end, like cursor’s UI, after all changes are.
3
1
16
RT @clefourrier: For all my eval friends wanting to re-run 3rd party evals, you'll need to BE CAREFUL WITH THE CHAT TEMPLATE: format/channe….
0
3
0
High quality, low latency. Exciting to see OpenAI’s open source models in the wild performing well.
Huge day for OSS models, . GPT-OSS-120b and 20b perform pretty well in @Stagehanddev,. Using @GroqInc for inference.
1
2
31
RT @zaidmukaddam: 1200 toks/s on @sciraai powered by @GroqInc!. use code `GPTOSS120` on checkout for 70% discount for first month on Scira….
0
9
0
OpenAI’s open source models running at unprecedented speed, low latency, and low price on @GroqInc . Not only that, we support built in browser search and code execution! It’s just a toggle! . Was very exciting putting this together for you all, keep feature requests coming!
2
4
42
RT @GroqInc: OpenAI’s open models are live and already running on Groq. Try gpt-oss-20B and gpt-oss-120B today. Groq delivers 128K context….
0
172
0
RT @nickscamara_: We're releasing an open source Lovable clone next week 👀. Paste any website URL and AI agents will instantly create a wor….
0
145
0
This is huge, one of the largest contributions we have made to open source. All led and driven by @AarushSah_ . It’s time to fix evals, open source.
1
4
35
RT @deansacoransky: We dropped five open source agent use cases: chat assistant, data enrichment, company research, crawl to RAG, meeting p….
tavily.com
Tavily is the real‑time search engine for AI agents and RAG workflows — Fast and secure APIs for web search and content extraction. Trusted by 600K+ developers.
0
7
0
RT @nickhaber: Super excited to see this work come out and be presented at @COLM_conf. This was a really delicate benchmarking task to get….
0
1
0
@GroqInc @Context7AI @upstash @vercel 4/ If you found the project useful, give it a star and repost!
Connect Kimi K2 to live docs for 30,000 popular frameworks 🔥. Most AI coding assistants, while impressive, have a significant gap — outdated implementation knowledge. I spend this weekend building a new open source project to fix that: Meet CodeWizard, Kimi K2 running fast on
0
0
4
@GroqInc @Context7AI @upstash @vercel 3/ The project is open to PRs/contributions! Would love to see other models supported, improved integration with context7 that leverages MCP get tools functionality, and other related MCP integrations!.
2
0
3
2/ CodeWizard uses Kimi K2 on @GroqInc (but can be altered to work with ANY OpenAI compatible model endpoint!), MCP connection to @Context7AI by @upstash, and nextjs on @vercel!.
1
0
5