EXAONE 4.0 32B (Reasoning) vs Apriel-v1.5-15B-Thinker
Comparing 2 AI models · 6 benchmarks · LG AI Research, ServiceNow
Most Affordable
Se
Apriel-v1.5-15B-Thinker
$0.00/1M
Highest Intelligence
LG
EXAONE 4.0 32B (Reasoning)
73.9% GPQA
Best for Coding
Se
Apriel-v1.5-15B-Thinker
39.2 Coding Index
Price Difference
Infinityx
input cost range
Composite Indices
Intelligence, Coding, Math
Standard Benchmarks
Academic and industry benchmarks
Benchmark Winners
6 tests
LG
EXAONE 4.0 32B (Reasoning)
4
- GPQA
- MMLU Pro
- LiveCodeBench
- MATH 500
Se
Apriel-v1.5-15B-Thinker
2
- HLE
- AIME 2025
| Metric | LG EXAONE 4.0 32B (Reasoning) | Se Apriel-v1.5-15B-Thinker |
|---|---|---|
| Pricing Per 1M tokens | ||
| Input Cost | $0.60/1M | $0.00/1M |
| Output Cost | $1.00/1M | $0.00/1M |
| Blended Cost 3:1 input/output ratio | $0.70/1M | — |
| Specifications | ||
| Organization Model creator | LG AI Research | ServiceNow |
| Release Date Launch date | Jul 15, 2025 | Sep 30, 2025 |
| Performance & Speed | ||
| Throughput Output speed | 93.6 tok/s | 146.0 tok/s |
| Time to First Token (TTFT) Initial response delay | 391ms | 164ms |
| Latency Time to first answer token | 21765ms | 13867ms |
| Composite Indices | ||
| Intelligence Index Overall reasoning capability | 42.6 | 51.6 |
| Coding Index Programming ability | 37.5 | 39.2 |
| Math Index Mathematical reasoning | 80.0 | 87.5 |
| Standard Benchmarks | ||
| GPQA Graduate-level reasoning | 73.9% | 71.3% |
| MMLU Pro Advanced knowledge | 81.8% | 77.3% |
| HLE Hard language evaluation | 10.5% | 12.0% |
| LiveCodeBench Real-world coding tasks | 74.7% | 72.8% |
| MATH 500 Mathematical problems | 97.7% | — |
| AIME 2025 Advanced math competition | 80.0% | 87.5% |
| AIME (Original) Math olympiad problems | 84.3% | — |
| SciCode Scientific code generation | 34.4% | 34.8% |
| LCR Code review capability | 14.0% | 20.0% |
| IFBench Instruction-following | 36.3% | 61.7% |
| TAU-bench v2 Tool use & agentic tasks | 17.3% | 68.4% |
| TerminalBench Hard CLI command generation | 3.5% | 9.9% |
Key Takeaways
Apriel-v1.5-15B-Thinker offers the best value at $0.00/1M, making it ideal for high-volume applications and cost-conscious projects.
EXAONE 4.0 32B (Reasoning) leads in reasoning capabilities with a 73.9% GPQA score, excelling at complex analytical tasks and problem-solving.
Apriel-v1.5-15B-Thinker achieves a 39.2 coding index, making it the top choice for software development and code generation tasks.
All models support context windows of ∞+ tokens, suitable for processing lengthy documents and maintaining extended conversations.