Model AI Terbaik untuk Menulis
Peringkat model bahasa berdasarkan kualitas tulisan. Diperbarui Agustus 2026. Berdasar data benchmark independen.
| # | Model | Penyedia | Skor | Kecepatan | Harga (output) |
|---|---|---|---|---|---|
| 1 | Claude Mythos Preview | Anthropic | 98.7 | 80 t/s | $15/M |
| 2 | Gemini 3.1 Pro | 78.0 | 136 t/s | $12/M | |
| 3 | Claude Fable 5.1 | Anthropic | 76.8 | 85 t/s | $15/M |
| 4 | Claude Mythos 5 | Anthropic | 75.8 | 50 t/s | $50/M |
| 5 | Claude Sonnet 4.6 (max) | Anthropic | 68.0 | 50 t/s | $15/M |
| 6 | Qwen3.8 Flash Next | Alibaba | 63.4 | 168 t/s | $0.4/M |
| 7 | Gemini 3 Pro | 60.6 | 120 t/s | $5/M | |
| 8 | GLM-5.3 Flash | Z.ai | 59.9 | 182 t/s | $0.3/M |
| 9 | Claude Fable 5.1 (XHigh) | Anthropic | 59.0 | 78 t/s | $15/M |
| 10 | Claude Fable 5.1 (High) | Anthropic | 59.0 | 92 t/s | $15/M |
| 11 | Claude Fable 5.1 (Medium) | Anthropic | 59.0 | 104 t/s | $15/M |
| 12 | Claude Fable 5.1 (Low) | Anthropic | 59.0 | 118 t/s | $15/M |
| 13 | Qwen3.7 Max | Alibaba | 58.6 | 203 t/s | $7.5/M |
| 14 | Claude Fable 5 | Anthropic | 58.4 | 71 t/s | $50/M |
| 15 | GPT-6 Astra (xhigh) | OpenAI | 58.0 | 76 t/s | $50/M |
| 16 | GPT-6 Astra (high) | OpenAI | 58.0 | 90 t/s | $50/M |
| 17 | Claude Opus 5.5 | Anthropic | 58.0 | 68 t/s | $25/M |
| 18 | GPT-6 Astra (medium) | OpenAI | 58.0 | 98 t/s | $50/M |
| 19 | GPT-6 Astra (low) | OpenAI | 58.0 | 110 t/s | $50/M |
| 20 | Gemini 3.8 Flash | 58.0 | 302 t/s | $3.75/M |
Claude Mythos Preview
Gemini 3.1 Pro
Qwen3.8 Flash Next
GLM-5.3 Flash
GPT-6 Astra (xhigh)Top 5 \u2014 Perbandingan cepat
Metodologi
Peringkat berdasar evaluasi benchmark independen. Skor dihitung dari beberapa benchmark terstandar dan dinormalisasi ke 0–100.