← All comparisons
DeepSeek R1 Distill Llama 70B vs Apriel-v1.5-15B-Thinker
DeepSeek vs ServiceNow — side-by-side benchmark comparison
| DeepSeek R1 Distill Llama 70B | Apriel-v1.5-15B-Thinker | |
|---|---|---|
| Intelligence Index | 16.0 | 28.3 |
| Coding Index | 11.4 | 18.7 |
| Math Index | 53.7 | 87.5 |
| Output speed (tok/s) | 46.8 | 0.0 |
| Blended price ($/1M) | $0.79 | $0.00 |
| Time to first token (s) | 0.33s | 0.00s |
| aime | 67.0% | — |
| aime 25 | 53.7% | 87.5% |
| artificial analysis coding index | 11.40 | 18.70 |
| artificial analysis intelligence index | 16.00 | 28.30 |
| artificial analysis math index | 53.70 | 87.50 |
| gpqa | 40.2% | 71.3% |
| hle | 6.1% | 12.0% |
| ifbench | 27.6% | 61.7% |
| lcr | 11.0% | 20.0% |
| livecodebench | 26.6% | 72.8% |
| math 500 | 93.5% | — |
| mmlu pro | 79.5% | 77.3% |
| scicode | 31.3% | 34.8% |
| tau2 | 21.9% | 68.4% |
| terminalbench hard | 1.5% | 10.6% |
Benchmark data from Artificial Analysis.