FunctionGemma
Ranked 2 of 4 tool-calling models by recommenders
A tiny open model for function calling that runs on-device and is praised for its local-first direction; critics say it is not reliable enough for tool calls.
A tiny open model for function calling that runs on-device and is praised for its local-first direction; critics say it is not reliable enough for tool calls.
13 people gave an opinion
- recommend it
- 11
- criticize it
- 2
- approval share
- 85%
Recommended in 3 of 13 months
- November 2025: not recommended
- December 2025: recommended
- January 2026: recommended
- February 2026: not recommended
- March 2026: not recommended
- April 2026: not recommended
- May 2026: not recommended
- June 2026: not recommended
- July 2026: not recommended
- August 2026: not recommended
- September 2026: recommended
- October 2026: not recommended
- November 2026: not recommended (month in progress)
Why it doesn’t carry the badge
- 11 people recommend itNeeds 10 or more, met.
- 5.5 recommenders for every criticNeeds 2 to 1 or better, met.
- Recommended in 3 of 13 monthsNeeds 4 or more, not met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Small, fast, runs on-device
“Check out the demo videos and colab that show you how to load this model onto a device relatively easily.”
Comment from 18 Dec 2025 in “FunctionGemma 270M Model” “Our goal here was the leanest weight and token base for blazing fast performance for you all.”
Comment from 19 Dec 2025 in “FunctionGemma 270M Model” “the parameters were megabytes and not gigabytes”
Comment from 19 Dec 2025 in “FunctionGemma 270M Model” - Comment from 18 Dec 2025 in “FunctionGemma 270M Model”
Praised local-first, on-device direction
“Much-appreciate the focus on local-first (on-device) !”
Comment from 18 Dec 2025 in “FunctionGemma 270M Model” “I felt that was a very under appreciated direction”
Comment from 19 Dec 2025 in “FunctionGemma 270M Model” “New models like functiongemma are promising”
Comment from 26 Dec 2025 in “Ask HN: What skills do you want to develop or improve in 2026?” - Comment from 19 Dec 2025 in “FunctionGemma 270M Model”
What people warn about
Each mark is one comment
Not reliable or capable enough
“it's still not quite capable enough for my target compared to FunctionGemma”
Comment from 18 Sep 2026 in “Show HN: Cactus Needle 3: 8-29MB automation models can match DeepSeek V4 Flash” “the problem is that is not reliable enough for a terminal assistant”
Comment from 4 Sep 2026 in “Show HN: TERMy – A fast terminal assistant that does not use LLMs” “I would be wary of having a LLM with 85% accuracy call tools on my system.”
Comment from 18 Dec 2025 in “FunctionGemma 270M Model” - Comment from 4 Sep 2026 in “Show HN: TERMy – A fast terminal assistant that does not use LLMs”
Other tool-calling models people recommend
The three most recommended other tool-calling models, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| Needle2 | 51 | 22 | 70% | 5 | — | Approved |
| FunctionGemmaThis page | 11 | 2 | 85% | 3 | — | — |
| Cactus Needle 3 | 10 | 12 | 45% | 1 | — | — |
| Needle 26M | 6 | 3 | 67% | 1 | — | — |
What people said
15 of the 16 comments that mention it, newest first. Each links to the original comment.
“it's still not quite capable enough for my target compared to FunctionGemma”
“FunctionGemma never worked well for me (without fine tuning).”
“the problem is that is not reliable enough for a terminal assistant”
“New models like functiongemma are promising”
“Our goal here was the leanest weight and token base for blazing fast performance for you all.”
“Its definitely a step in that direction”
“I felt that was a very under appreciated direction”
“the parameters were megabytes and not gigabytes”
“I would be wary of having a LLM with 85% accuracy call tools on my system.”
“Much-appreciate the focus on local-first (on-device) !”
“Something like this would be a good use case for the model.”
“Check out the demo videos and colab that show you how to load this model onto a device relatively easily.”
“It's been quite effective my testing, and the finetuning guides should show you how to add in your own capabilities.”
“Very cool model! Congrats on the work!”
“Amazing it can run in the browser.”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.