모델 정가표는 어디에나 있습니다. 없는 것은 «에이전트 루프에서 실제로 얼마가 나가는가» 입니다. 반복 컨텍스트가 원가의 대부분을 차지하는 실사용에서는 프롬프트 캐시가 먹는지에 따라 같은 모델이 5~20배 갈립니다. 우리는 그것을 직접 재서 여기에 공개합니다.
같은 프리픽스로 예열 1회 + 반복 N회를 호출하고, 반복 중 최저 비용 회차를 정상 상태로 봅니다. 캐시는 되거나 안 되거나가 아니라 콜드·웜이 섞이므로 «N회 중 몇 번 걸렸나»를 함께 표기합니다. 실효 입력단가 = 정상 상태 비용 ÷ 그 회차 입력토큰 × 100만.
회차원가 환산 기준(실사용 프로파일): 입력 484 · 출력 13,570 · 캐시읽기 3,072,894 · 평균 34.4턴
| 모델 | 컨텍스트 | 정가 입력 /M | 정가 출력 /M | 캐시 선언 | 캐시 실측 | 실효 입력단가 /M | 회차원가 환산 |
|---|---|---|---|---|---|---|---|
| deepseek/deepseek-v4-flash | 1M | $0.140 | $0.280 | 있음 | 먹음2/3 | $0.028-80% | $0.105 |
| qwen/qwen3-coder-flash | 1M | $0.195 | $0.975 | 있음 | 먹음2/3 | $0.043-78% | $0.154 |
| qwen/qwen3-coder-plus | 1M | $0.650 | $3.250 | 있음 | 먹음3/3 | $0.144-78% | $0.515 |
| moonshotai/kimi-k3 | 1M | $3.000 | $15.000 | 있음 | 먹음2/3 | $0.769-74% | $1.454 |
| 모델 | 컨텍스트 | 정가 입력 /M | 정가 출력 /M | 캐시 선언 |
|---|---|---|---|---|
| cohere/north-mini-code:free | 256K | $0.000 | $0.000 | — |
| google/gemma-4-26b-a4b-it:free | 262K | $0.000 | $0.000 | — |
| google/gemma-4-31b-it:free | 262K | $0.000 | $0.000 | — |
| google/lyria-3-clip-preview | 1M | $0.000 | $0.000 | — |
| google/lyria-3-pro-preview | 1M | $0.000 | $0.000 | — |
| inclusionai/ling-3.0-tiny:free | 262K | $0.000 | $0.000 | — |
| nvidia/nemotron-3.5-content-safety:free | 128K | $0.000 | $0.000 | — |
| nvidia/nemotron-3.5-lightning:free | 1M | $0.000 | $0.000 | — |
| nvidia/nemotron-3-nano-30b-a3b:free | 256K | $0.000 | $0.000 | — |
| nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free | 256K | $0.000 | $0.000 | — |
| nvidia/nemotron-3-super-120b-a12b:free | 262K | $0.000 | $0.000 | — |
| nvidia/nemotron-3-ultra-550b-a55b:free | 1M | $0.000 | $0.000 | — |
| nvidia/nemotron-nano-12b-v2-vl:free | 128K | $0.000 | $0.000 | — |
| nvidia/nemotron-nano-9b-v2:free | 128K | $0.000 | $0.000 | — |
| openai/gpt-oss-20b:free | 131K | $0.000 | $0.000 | — |
| openrouter/free | 200K | $0.000 | $0.000 | — |
| poolside/laguna-s-2.1:free | 262K | $0.000 | $0.000 | — |
| poolside/laguna-xs-2.1:free | 262K | $0.000 | $0.000 | — |
| inclusionai/ling-2.6-flash | 262K | $0.010 | $0.030 | 있음 |
| ibm-granite/granite-4.0-h-micro | 131K | $0.017 | $0.112 | — |
| mistralai/mistral-nemo | 131K | $0.019 | $0.030 | — |
| inclusionai/ling-3.0-flash | 262K | $0.021 | $0.063 | 있음 |
| nex-agi/nex-n2-mini | 262K | $0.025 | $0.100 | 있음 |
| openai/gpt-5-nano:batch | 400K | $0.025 | $0.200 | 있음 |
| meta-llama/llama-3.2-1b-instruct | 60K | $0.027 | $0.201 | — |
| openai/gpt-oss-20b | 131K | $0.030 | $0.130 | 있음 |
| qwen/qwen3.7-flash | 1M | $0.030 | $0.130 | 있음 |
| upstage/solar-pro4 | 524K | $0.030 | $0.120 | 있음 |
| amazon/nova-micro-v1 | 128K | $0.035 | $0.140 | — |
| openai/gpt-oss-120b | 131K | $0.037 | $0.170 | — |
| cohere/command-r7b-12-2024 | 128K | $0.037 | $0.150 | — |
| sao10k/l3-lunaris-8b | 8K | $0.040 | $0.050 | — |
| qwen/qwen3-30b-a3b-instruct-2507 | 262K | $0.048 | $0.193 | — |
| google/gemini-2.5-flash-lite:batch | 1M | $0.050 | $0.200 | 있음 |
| google/gemma-3-12b-it | 131K | $0.050 | $0.150 | — |
| google/gemma-3-4b-it | 131K | $0.050 | $0.100 | — |
| ibm-granite/granite-4.1-8b | 131K | $0.050 | $0.100 | 있음 |
| meta-llama/llama-3.1-8b-instruct | 131K | $0.050 | $0.080 | 있음 |
| meta-llama/llama-3.2-3b-instruct | 131K | $0.050 | $0.330 | — |
| mistralai/mistral-small-24b-instruct-2501 | 33K | $0.050 | $0.080 | — |
| nvidia/nemotron-3-nano-30b-a3b | 262K | $0.050 | $0.200 | 있음 |
| openai/gpt-4.1-nano:batch | 1M | $0.050 | $0.200 | 있음 |
| openai/gpt-5-nano | 400K | $0.050 | $0.400 | 있음 |
| amazon/nova-lite-v1 | 300K | $0.060 | $0.240 | — |
| google/gemma-3n-e4b-it | 33K | $0.060 | $0.120 | — |
| gryphe/mythomax-l2-13b | 8K | $0.060 | $0.060 | — |
| poolside/laguna-xs-2.1 | 262K | $0.060 | $0.120 | 있음 |
| z-ai/glm-4.7-flash | 203K | $0.060 | $0.400 | 있음 |
| tencent/hy3-preview | 262K | $0.063 | $0.210 | 있음 |
| qwen/qwen3.5-flash-02-23 | 1M | $0.065 | $0.260 | — |
| microsoft/phi-4 | 16K | $0.070 | $0.140 | — |
| qwen/qwen3-coder-30b-a3b-instruct | 262K | $0.070 | $0.280 | — |
| bytedance-seed/seed-1.6-flash | 262K | $0.075 | $0.300 | — |
| inclusionai/ling-2.6-1t | 262K | $0.075 | $0.625 | 있음 |
| inclusionai/ring-2.6-1t | 262K | $0.075 | $0.625 | 있음 |
| openai/gpt-4o-mini:batch | 128K | $0.075 | $0.300 | 있음 |
| openai/gpt-oss-safeguard-20b | 131K | $0.075 | $0.300 | 있음 |
| ~deepseek/deepseek-v4-flash-latest | 1M | $0.080 | $0.252 | 있음 |
| deepseek/deepseek-v4-flash-0731 | 1M | $0.080 | $0.180 | 있음 |
| google/gemma-3-27b-it | 262K | $0.080 | $0.450 | 있음 |
카탈로그 최종 동기화: 2026-08-11 13:16 UTC