Kiln AI
@getkiln
Followers
11
Following
4
Media
2
Statuses
17
The easiest way to build AI products. Our open tooling helps you build evals, synthetic data, and evals.
Joined May 2025
Kiln v0.19 just came out with a new model library, GPT-5, GPT OSS, GLM, Qwen 2507, Cerebras support, SiliconFlow support, and much more!. Download it now:
github.com
What's Changed Model Library: Browse all available models and their supported capabilities in our new model library. New AI Providers: Cerebras, SiliconFlow.cn New Models: GPT-5, GPT OSS, GLM,...
0
1
2
New Models: This release includes a range of new models including Kimi K2, Grok 4, Deepseek R1 0528, and Llama 4! Grok 4 has been fun for generating adversarial synthetic data.
1
0
1
Introducing 'Kiln Issues': Software teams have tools like Github and Jira issue tracking. What should AI teams use?.Kiln Issues are a AI-native issue tracker. It doesn’t just collect bug reports, but keeps the structured data needed to reproduce, evaluate and fix AI systems.
1
0
1
Synthetic Data V2: adds more use cases, including fine-tuning, generating eval data, issue resolution, and more!.
1
0
1
Evals V2: we've simplified eval creation, improved prompts, eval templates, and added a new "compare" view to find the best model for your AI workload.
1
0
1
We just shipped the biggest Kiln release ever! In includes evals v2, synthetic data v2, and a a whole new concept: Kiln Issues!. Read more:
1
1
1
New Blog Post: Many Small Evals Beat One Big Eval, Every Time. Small evals make evals much more maintainable, rapid and iterative. Maybe most importantly: your team will actually enjoy using them!.
0
1
3
RT @bentlegen: Big thanks to everyone who made it out to last night's @devtoolsTO event. We got to see the first in-person premieres of so….
0
7
0
RT @bentlegen: Only 2 weeks away! Demos from @SST_dev @getkiln @vltpkg @DevCycleHQ . Sponsored by @OpsLevelHQ @getsentry and @bdc_ca.
0
7
0
Want to know what makes a fine-tuned model great — or just lucky?.We ran ablations on Google’s Gemma 3 to show how prompt quality, data mix, and quantity affect performance. A practical guide for building better LLMs:.🧪
0
1
1
Why fine-tune an LLM when you can just prompt it?.Turns out: speed, cost, control — and better UX. We break down when fine-tuning is worth it and how to get started fast.
1
1
1
Kiln makes synthetic data easy. Intuitive interactive UI, human guidance, and deep integrations for fine-tuning and evals. Get started at
0
1
1
Kiln v0.16 was released! It includes Qwen 3, Gemma 3, Evals Guides, and Fine Tuning Updates. Learn more here:
0
1
1