AI BENCHY Compare
Model yang Dibandingkan
Bandingkan:
Benchmark dihasilkan dari suite pengujian AI BENCHY pada: 2026-03-03
| Metrik | OpenAI: GPT-5.2 Chat none Rilis: 2025-12-11 | OpenAI: GPT-5.3 Chat none Rilis: 2026-03-03 | Google: Gemini 3.1 Flash Lite Preview low Rilis: 2026-03-03 |
|---|---|---|---|
| Peringkat | #12 | #14 | #8 |
| Skor Rata-rata | 7.41 | 7.27 | 7.84 |
| Konsistensi | 9.45 | 8.26 | 10.00 |
| Biaya per hasil | 2.261 | 2.835 | 0.151 |
| Total Biaya | $0.227 | $0.256 | $0.016 |
| Tes benar | |||
| Tingkat lulus per percobaan | 73.8% | 73.8% | 71.4% |
| Tes tidak stabil | 1 | 3 | 0 |
| Token output | 14,267 | 16,339 | 1,317 |
| Token penalaran | 0 | 0 | 6,126 |
Skor vs Total Biaya
Rincian Kategori
| Trik anti-AI | Skor | Konsistensi | Tingkat lulus per percobaan | Tes tidak stabil | Tes benar | Token output | Token penalaran |
|---|---|---|---|---|---|---|---|
| OpenAI: GPT-5.2 Chat | 10.00 | 10.00 | 100.0% | 0 | 1,651 | 0 | |
| OpenAI: GPT-5.3 Chat | 7.33 | 7.49 | 77.8% | 1 | 3,091 | 0 | |
| Google: Gemini 3.1 Flash Lite Preview | 7.00 | 10.00 | 66.7% | 0 | 456 | 1,224 |
| Parsing dan ekstraksi data | Skor | Konsistensi | Tingkat lulus per percobaan | Tes tidak stabil | Tes benar | Token output | Token penalaran |
|---|---|---|---|---|---|---|---|
| OpenAI: GPT-5.2 Chat | 9.88 | 10.00 | 100.0% | 0 | 980 | 0 | |
| OpenAI: GPT-5.3 Chat | 9.88 | 10.00 | 100.0% | 0 | 942 | 0 | |
| Google: Gemini 3.1 Flash Lite Preview | 9.88 | 10.00 | 100.0% | 0 | 291 | 696 |
| Spesifik domain | Skor | Konsistensi | Tingkat lulus per percobaan | Tes tidak stabil | Tes benar | Token output | Token penalaran |
|---|---|---|---|---|---|---|---|
| OpenAI: GPT-5.2 Chat | 4.00 | 10.00 | 33.3% | 0 | 7,810 | 0 | |
| OpenAI: GPT-5.3 Chat | 1.00 | 4.41 | 33.3% | 2 | 8,264 | 0 | |
| Google: Gemini 3.1 Flash Lite Preview | 4.00 | 10.00 | 33.3% | 0 | 18 | 1,212 |
| Kepatuhan instruksi | Skor | Konsistensi | Tingkat lulus per percobaan | Tes tidak stabil | Tes benar | Token output | Token penalaran |
|---|---|---|---|---|---|---|---|
| OpenAI: GPT-5.2 Chat | 5.50 | 6.13 | 66.7% | 1 | 1,528 | 0 | |
| OpenAI: GPT-5.3 Chat | 8.50 | 9.99 | 50.0% | 0 | 1,455 | 0 | |
| Google: Gemini 3.1 Flash Lite Preview | 8.50 | 10.00 | 50.0% | 0 | 72 | 753 |
| Puzzle Solving | Skor | Konsistensi | Tingkat lulus per percobaan | Tes tidak stabil | Tes benar | Token output | Token penalaran |
|---|---|---|---|---|---|---|---|
| OpenAI: GPT-5.2 Chat | 7.00 | 10.00 | 66.7% | 0 | 1,743 | 0 | |
| OpenAI: GPT-5.3 Chat | 10.00 | 10.00 | 100.0% | 0 | 1,726 | 0 | |
| Google: Gemini 3.1 Flash Lite Preview | 10.00 | 10.00 | 100.0% | 0 | 243 | 1,248 |
| Pemanggilan alat | Skor | Konsistensi | Tingkat lulus per percobaan | Tes tidak stabil | Tes benar | Token output | Token penalaran |
|---|---|---|---|---|---|---|---|
| OpenAI: GPT-5.2 Chat | 10.00 | 10.00 | 100.0% | 0 | 555 | 0 | |
| OpenAI: GPT-5.3 Chat | 10.00 | 10.00 | 100.0% | 0 | 861 | 0 | |
| Google: Gemini 3.1 Flash Lite Preview | 10.00 | 10.00 | 100.0% | 0 | 237 | 993 |
Perbandingan Cepat
Ganti Pasangan Perbandingan
GPT-5.3 ChatnonevsGLM 5mediumClaude Sonnet 4.6mediumvsGPT-5.2 ChatnoneGemini 3.1 Flash Lite PreviewlowvsGPT-5.3-CodexmediumGPT-5.2 ChatnonevsGLM 5mediumClaude Sonnet 4.6mediumvsGPT-5.3 ChatnoneGemini 2.5 FlashmediumvsGPT-5.3 ChatnoneGPT-5.3 ChatnonevsStep 3.5 FlashmediumTersedia gratisDeepSeek V3.2mediumvsGPT-5.3 ChatnoneGemini 2.5 FlashmediumvsGPT-5.2 ChatnoneGemini 3.1 Flash Lite PreviewhighvsGPT-5.2 ChatnoneGemini 3.1 Flash Lite PreviewmediumvsGPT-5.3 ChatnoneClaude Sonnet 4.6mediumvsGemini 3.1 Flash Lite Previewlow