Best model for tool use
Models ranked by their tool use score, 0 to 100, as on the category page. Only ranked models with enough scored items are included, and models are grouped by total parameters.
Read the memory figure as an estimate, not a guarantee. The memory a model needs is estimated from its file: the weights plus about 20% for the context of a single conversation. It is not a measurement, and a long conversation or several at once needs more. A model with no file size on record is not counted as fitting any amount of memory. Speeds, and the GPU memory shown on the speed page, were measured on this benchmark's server, not on yours. That server kept 1 to 8 answers in flight at a time and loaded a context window of 16,384 tokens for most models; 42 of the models ranked ran at a shorter context. Speeds were measured on one laptop and are comparable only between models on the same backend. See the settings and the memory method.
All sizes
Aya Expanse 8B leads here with 100.0 on a scale of 0 to 100; its Overall is 72.5 (95% interval 67.5–77.3).
| Place | Model | Tool use (0 to 100) | Overall | Rank range | Decode (tok/s) | Memory needed (GB, est.) | Size class |
|---|---|---|---|---|---|---|---|
| 1 | Aya Expanse 8B | 100.020 of 20 items scored | 72.567.5–77.3 | 90–154 | 10.7ROCm 7.1 | 6.1 | 4–9B |
| 1 | Cogito 14B V1 Preview Qwenfine-tune of Qwen 2.5 | 100.020 of 20 items scored | 90.386.8–93.4 | 11–79 | 6.6ROCm 7.1 | 10.8 | 9–16B |
| 1 | Cogito 8B V1 Preview Llamafine-tune of Llama 3.1 | 100.020 of 20 items scored | 83.979.6–87.8 | 47–105 | 11.2ROCm 7.1 | 5.9 | 4–9B |
| 1 | Deepcoder 14B Previewfine-tune of DeepSeek-R1-Distill-Qwen 14Bcode model | 100.020 of 20 items scored | 74.569.5–79.0 | 84–146 | 6.7ROCm 7.1 | 10.8 | 9–16B |
| 1 | Deepseek R1 8B Llama Distill | 100.020 of 20 items scored | 81.776.9–86.1 | 56–122 | 11.9ROCm 7.1 | 5.9 | 4–9B |
Tiny (under 3B)
Qwen3 1.7B leads its size class with 100.0 on a scale of 0 to 100; its Overall is 76.0 (95% interval 71.3–80.9).
| Place | Model | Tool use (0 to 100) | Overall | Rank range | Decode (tok/s) | Memory needed (GB, est.) | Size class |
|---|---|---|---|---|---|---|---|
| 1 | Qwen3 1.7B | 100.020 of 20 items scored | 76.071.3–80.9 | 77–141 | 38.5ROCm 7.1 | 1.7 | Tiny (under 3B) |
| 1 | Qwen3 VL 2B Instruct | 100.020 of 20 items scored | 77.472.5–82.1 | 68–137 | 26.5ROCm 7.1 | 2.3 | Tiny (under 3B) |
| 3 | Lfm2.5 Thinking 1.2B | 97.520 of 20 items scored | 74.369.5–79.1 | 84–148 | 205.0ROCm 7.1 | 0.9 | Tiny (under 3B) |
| 4 | Exaone3.5 2.4B Instruct | 95.020 of 20 items scored | 76.370.9–81.3 | 73–141 | 100.9ROCm 7.1 | 1.9 | Tiny (under 3B) |
| 4 | Granite3.2 2B Instruct | 95.020 of 20 items scored | 63.858.7–69.0 | 132–172 | 30.1ROCm 7.1 | 1.8 | Tiny (under 3B) |
3–4B
Granite4 Micro H leads its size class with 100.0 on a scale of 0 to 100; its Overall is 76.6 (95% interval 72.0–81.4).
| Place | Model | Tool use (0 to 100) | Overall | Rank range | Decode (tok/s) | Memory needed (GB, est.) | Size class |
|---|---|---|---|---|---|---|---|
| 1 | Granite4 Micro H | 100.020 of 20 items scored | 76.672.0–81.4 | 72–139 | 13.4ROCm 7.1 | 2.3 | 3–4B |
| 2 | Granite4.2 3B | 95.020 of 20 items scored | 78.273.7–83.1 | 72–140 | 24.3ROCm 7.1 | 2.7 | 3–4B |
| 2 | Ministral 3 3B Instruct 2512 | 95.020 of 20 items scored | 80.976.5–85.2 | 61–124 | 70.9ROCm 7.1 | 3.6 | 3–4B |
| 2 | Phi-4-mini-instruct | 95.020 of 20 items scored | 81.677.1–86.0 | 62–123 | 17.1ROCm 7.1 | 3.0 | 3–4B |
| 2 | Phi3 3.8B Mini 128K Instruct | 95.020 of 20 items scored | 62.958.0–67.7 | 137–176 | 63.9ROCm 7.1 | 2.9 | 3–4B |
4–9B
Aya Expanse 8B leads its size class with 100.0 on a scale of 0 to 100; its Overall is 72.5 (95% interval 67.5–77.3).
| Place | Model | Tool use (0 to 100) | Overall | Rank range | Decode (tok/s) | Memory needed (GB, est.) | Size class |
|---|---|---|---|---|---|---|---|
| 1 | Aya Expanse 8B | 100.020 of 20 items scored | 72.567.5–77.3 | 90–154 | 10.7ROCm 7.1 | 6.1 | 4–9B |
| 1 | Cogito 8B V1 Preview Llamafine-tune of Llama 3.1 | 100.020 of 20 items scored | 83.979.6–87.8 | 47–105 | 11.2ROCm 7.1 | 5.9 | 4–9B |
| 1 | Deepseek R1 8B Llama Distill | 100.020 of 20 items scored | 81.776.9–86.1 | 56–122 | 11.9ROCm 7.1 | 5.9 | 4–9B |
| 1 | DeepSeek-R1-Distill-Qwen-7B | 100.020 of 20 items scored | 73.668.9–78.3 | 82–155 | 13.8ROCm 7.1 | 5.6 | 4–9B |
| 1 | EXAONE-3.5-7.8B-Instruct | 100.020 of 20 items scored | 82.077.6–86.4 | 59–122 | 9.5ROCm 7.1 | 5.7 | 4–9B |
9–16B
Cogito 14B V1 Preview Qwen leads its size class with 100.0 on a scale of 0 to 100; its Overall is 90.3 (95% interval 86.8–93.4).
| Place | Model | Tool use (0 to 100) | Overall | Rank range | Decode (tok/s) | Memory needed (GB, est.) | Size class |
|---|---|---|---|---|---|---|---|
| 1 | Cogito 14B V1 Preview Qwenfine-tune of Qwen 2.5 | 100.020 of 20 items scored | 90.386.8–93.4 | 11–79 | 6.6ROCm 7.1 | 10.8 | 9–16B |
| 1 | Deepcoder 14B Previewfine-tune of DeepSeek-R1-Distill-Qwen 14Bcode model | 100.020 of 20 items scored | 74.569.5–79.0 | 84–146 | 6.7ROCm 7.1 | 10.8 | 9–16B |
| 1 | DeepSeek-R1-Distill-Qwen-14B | 100.020 of 20 items scored | 85.381.2–89.2 | 38–106 | 7.1ROCm 7.1 | 10.8 | 9–16B |
| 1 | Gemma 4 12B IT | 100.020 of 20 items scored | 98.697.0–99.8 | 1–12 | 8.1ROCm 7.1 | 8.9 | 9–16B |
| 1 | gemma-3-12b-it | 100.020 of 20 items scored | 92.789.7–95.4 | 8–66 | 6.1ROCm 7.1 | 9.8 | 9–16B |
16–35B
Devstral 24B Small 2505 leads its size class with 100.0 on a scale of 0 to 100; its Overall is 88.9 (95% interval 85.6–92.0).
| Place | Model | Tool use (0 to 100) | Overall | Rank range | Decode (tok/s) | Memory needed (GB, est.) | Size class |
|---|---|---|---|---|---|---|---|
| 1 | Devstral 24B Small 2505code model | 100.020 of 20 items scored | 88.985.6–92.0 | 23–86 | 4.2ROCm 7.1 | 17.2 | 16–35B |
| 1 | Devstral Small 2 24B Instruct (2512)code model | 100.020 of 20 items scored | 92.189.3–94.7 | 9–67 | 3.1ROCm 7.1 | 18.2 | 16–35B |
| 1 | Exaone3.5 32B Instruct | 100.020 of 20 items scored | 85.080.9–88.8 | 46–106 | 3.1ROCm 7.1 | 23.2 | 16–35B |
| 1 | gemma-3-27b-it | 100.020 of 20 items scored | 95.893.7–97.5 | 3–36 | 2.7ROCm 7.1 | 20.9 | 16–35B |
| 1 | gemma-4-26B-A4B-it | 100.020 of 20 items scored | 97.895.8–99.3 | 1–22 | 15.7ROCm 7.1 | 21.6 | 16–35B |
> 35B
GLM-4.5-Air leads its size class with 100.0 on a scale of 0 to 100; its Overall is 92.8 (95% interval 89.7–95.6).
| Place | Model | Tool use (0 to 100) | Overall | Rank range | Decode (tok/s) | Memory needed (GB, est.) | Size class |
|---|---|---|---|---|---|---|---|
| 1 | GLM-4.5-Air | 100.020 of 20 items scored | 92.889.7–95.6 | 6–68 | 6.0Vulkan | 87.6 | > 35B |
| 1 | gpt-oss-120b | 100.020 of 20 items scored | 98.696.8–100.0 | 1–14 | 11.0Vulkan | 78.4 | > 35B |
| 1 | Llama-3.3-70B-Instruct | 100.020 of 20 items scored | 94.391.3–96.7 | 4–63 | 1.4Vulkan | 51.0 | > 35B |
| 1 | Qwen3-Coder-Nextcode model | 100.020 of 20 items scored | 96.494.2–98.3 | 2–42 | 18.1Vulkan | 62.1 | > 35B |
| 1 | Sage-A1.5-Q4_K_M | 100.020 of 20 items scored | 94.591.8–96.7 | 5–56 | 22.6ROCm 7.1 | 26.4 | > 35B |
A higher place is not always a clear win. When two models have overlapping rank ranges, this battery cannot tell them apart; see how sure a score is. A dash means not measured or not scored, which is different from 0. Speeds were measured on one laptop and are comparable only between models on the same backend.
More: all the best-model pages, the hardware picker, speed and memory for every model.