Model AI Terbaik untuk Agen Koding
Peringkat model bahasa berdasarkan performa pengkodean agentik. Diperbarui Agustus 2026. Berdasar data benchmark independen.
| # | Model | Penyedia | Skor | Kecepatan | Harga (output) |
|---|---|---|---|---|---|
| 1 | Claude Fable 5.1 | Anthropic | 94.6 | 85 t/s | $15/M |
| 2 | Claude Fable 5 | Anthropic | 91.4 | 71 t/s | $50/M |
| 3 | Kimi K3 | Moonshot AI | 84.6 | 37 t/s | $15/M |
| 4 | Claude Fable 5.1 (XHigh) | Anthropic | 83.2 | 78 t/s | $15/M |
| 5 | Claude Fable 5.1 (High) | Anthropic | 83.2 | 92 t/s | $15/M |
| 6 | Claude Fable 5.1 (Medium) | Anthropic | 83.2 | 104 t/s | $15/M |
| 7 | Claude Fable 5.1 (Low) | Anthropic | 83.2 | 118 t/s | $15/M |
| 8 | Claude Opus 5.5 | Anthropic | 82.0 | 68 t/s | $25/M |
| 9 | Claude Opus 5.5 (XHigh) | Anthropic | 72.0 | 62 t/s | $25/M |
| 10 | Claude Opus 5.5 (High) | Anthropic | 72.0 | 74 t/s | $25/M |
| 11 | Claude Opus 5.5 (Medium) | Anthropic | 72.0 | 82 t/s | $25/M |
| 12 | Claude Opus 5.5 (Low) | Anthropic | 72.0 | 92 t/s | $25/M |
| 13 | GPT-6 Astra | OpenAI | 70.4 | 82 t/s | $50/M |
| 14 | Claude Opus 4.8 | Anthropic | 68.1 | 50 t/s | $15/M |
| 15 | Gemini 3.5 Flash | 65.4 | 267 t/s | $9/M | |
| 16 | GPT-6 Astra (xhigh) | OpenAI | 64.8 | 76 t/s | $50/M |
| 17 | GPT-6 Astra (high) | OpenAI | 64.8 | 90 t/s | $50/M |
| 18 | GPT-6 Astra (medium) | OpenAI | 64.8 | 98 t/s | $50/M |
| 19 | GPT-6 Astra (low) | OpenAI | 64.8 | 110 t/s | $50/M |
| 20 | GPT-6 Luna | OpenAI | 64.4 | 120 t/s | $20/M |
Claude Fable 5.1
Kimi K3
GPT-6 Astra
Gemini 3.5 FlashTop 5 \u2014 Perbandingan cepat
Metodologi
Peringkat berdasar evaluasi benchmark independen. Skor dihitung dari beberapa benchmark terstandar dan dinormalisasi ke 0–100.