Input/output/cache billed separately. Model unit prices follow each model provider's official pricing.
Balance never expires
OpenAI official pricing vs 15code pricing. GPT-6 Luna is for low-cost, fast tasks; GPT-6.1 Sol is for high-quality tasks.
| Model | Official Input | 15code Input | Official Cache | 15code Cache | Official Output | 15code Output | Savings |
|---|---|---|---|---|---|---|---|
| GPT-6 Luna | $0.10/M | $0.10/M | $0.01/M | $0.01/M | $0.50/M | $0.50/M | Same · low-cost and fast |
| GPT-6.1 Sol · standard | $2/M | $2/M | $0.10/M | $0.10/M | $10/M | $10/M | Same |
| GPT-6.1 Sol · long context | $4/M | $4/M | $0.20/M | $0.20/M | $15/M | $15/M | Same |
| GPT-6 Sol | $2/M | $2/M | $0.20/M | $0.20/M | $10/M | $10/M | Same · high quality |
| GPT-5.6 Terra | $2/M | $2/M | $0.20/M | $0.20/M | $12/M | $12/M | Same · default |
| GPT-5.6 Sol | $5/M | $5/M | $0.5/M | $0.5/M | $30/M | $30/M | Same · high quality |
GPT-6.1 Sol (gpt-6.1-sol): USD per million tokens. For total input (including cache) ≤ 272,000: input $2, cache read $0.10, cache write $2.50, output $10. For total input > 272,000, the whole request uses long-context rates: input $4, cache read $0.20, cache write $5, output $15. Context: 1.05M; maximum output: 128K. Use the Responses API for tool calls; Chat Completions supports plain chat without tools.
Anthropic official pricing vs 15code pricing (input/output comparison)
| Model | Official Input | 15code Input | Official Cache | 15code Cache | Official Output | 15code Output | Savings |
|---|---|---|---|---|---|---|---|
| Claude Opus 5.0 | $5/M | $5/M | $0.5/M | $0.5/M | $25/M | $25/M | Same |
| Claude Sonnet 5 | $2/M | $2/M | $0.20/M | $0.20/M | $10/M | $10/M | Same · promo through 2026-08-31 |
Each billing component uses the lower of the official benchmark and the previous 15code price. Qwen3.8 is a local unrestricted quantized model and keeps the previous Qwen3.6 pricing.
| Model | Official Input | 15code Input | Official Cache | 15code Cache | Official Output | 15code Output | Savings |
|---|---|---|---|---|---|---|---|
| GLM-5.2 | $1.16/M | $1/M | $0.26/M | $0.26/M | $4.06/M | $3/M | Lowest per component |
| Kimi K3 | $3/M | $3/M | $0.30/M | $0.30/M | $15/M | $15/M | Official USD price · live |
| DeepSeek V4 Pro | $0.66/M | $0.66/M | $0.022/M | $0.022/M | $1.98/M | $1.98/M | Official off-peak USD price · live |
| Qwen3.8 27B Unrestricted | $0.261/M | $0.20/M | — | $0.10/M | $1.565/M | $0.50/M | Special price · local quantized |
Tiered bonus stacks on top of base amount; first recharge gets an extra +5%; balance never expires
| You Pay | Bonus | You Get | Credits | Good for |
|---|---|---|---|---|
| $1 | — | $1 | 1,000,000 | Trial (min) |
| $10 | +10% | $11.00 | 11,000,000 | Light usage |
| $30 Popular | +20% | $36.00 | 36,000,000 | Daily dev |
| $80 | +30% | $104.00 | 104,000,000 | Pro users |
| $150 Best | +40% | $210.00 | 210,000,000 | Team/Enterprise |
💡 Extra +5% on your first recharge · Refer friends, earn 5% of their first recharge · 1 USD = 1,000,000 credits