renjianbo
fe37f8be5c
- Backend: token-level streaming via on_delta callback + asyncio.Queue bridge, answer_chunk SSE events; final event carries token_usage with cost_yuan; kimi pricing in cost_estimator - Frontend: typewriter rendering of answer chunks; token/cost in message meta; abort marks message '已停止' and no longer triggers duplicate non-stream fallback; session pin/delete in dropdown (pinned first); date dividers across days
81 lines
2.3 KiB
Python
81 lines
2.3 KiB
Python
"""LLM 成本估算 — 基于模型定价和 Token 用量估算费用"""
|
||
from __future__ import annotations
|
||
|
||
import logging
|
||
from typing import Dict, Optional, Tuple
|
||
|
||
logger = logging.getLogger(__name__)
|
||
|
||
# 模型定价 (per 1M tokens, USD) — 2024 Q4 参考值
|
||
MODEL_PRICING: Dict[str, Tuple[float, float]] = {
|
||
# (input_price_per_1M, output_price_per_1M)
|
||
"gpt-4o": (2.50, 10.00),
|
||
"gpt-4o-mini": (0.15, 0.60),
|
||
"gpt-4-turbo": (10.00, 30.00),
|
||
"gpt-3.5-turbo": (0.50, 1.50),
|
||
"deepseek-chat": (0.14, 0.28),
|
||
"deepseek-v3": (0.27, 1.10),
|
||
"deepseek-r1": (0.55, 2.19),
|
||
"deepseek-v4-flash": (0.14, 0.28),
|
||
"deepseek-v4-pro": (0.27, 1.10),
|
||
"claude-3-opus": (15.00, 75.00),
|
||
"claude-3-sonnet": (3.00, 15.00),
|
||
"claude-3-haiku": (0.25, 1.25),
|
||
"claude-3.5-sonnet": (3.00, 15.00),
|
||
"claude-3.5-haiku": (1.00, 5.00),
|
||
"kimi-k3": (3.00, 15.00),
|
||
"kimi-k2": (0.56, 2.22),
|
||
}
|
||
|
||
# 缓存未命中模型的默认定价
|
||
DEFAULT_PRICING = (1.00, 4.00)
|
||
|
||
|
||
def estimate_cost(
|
||
model: str,
|
||
prompt_tokens: int = 0,
|
||
completion_tokens: int = 0,
|
||
) -> float:
|
||
"""估算单次 LLM 调用的费用（USD）。
|
||
|
||
Args:
|
||
model: 模型名称（模糊匹配）
|
||
prompt_tokens: 输入 token 数
|
||
completion_tokens: 输出 token 数
|
||
"""
|
||
input_price, output_price = _get_price(model)
|
||
cost = (prompt_tokens / 1_000_000) * input_price + (
|
||
completion_tokens / 1_000_000
|
||
) * output_price
|
||
return round(cost, 6)
|
||
|
||
|
||
def estimate_cost_yuan(
|
||
model: str,
|
||
prompt_tokens: int = 0,
|
||
completion_tokens: int = 0,
|
||
exchange_rate: float = 7.2,
|
||
) -> float:
|
||
"""估算费用（人民币）。"""
|
||
return round(estimate_cost(model, prompt_tokens, completion_tokens) * exchange_rate, 4)
|
||
|
||
|
||
def _get_price(model: str) -> Tuple[float, float]:
|
||
"""模糊匹配模型定价。"""
|
||
model_lower = model.lower().replace("-", "").replace(".", "")
|
||
for key, price in MODEL_PRICING.items():
|
||
if key.replace("-", "").replace(".", "") in model_lower:
|
||
return price
|
||
return DEFAULT_PRICING
|
||
|
||
|
||
def get_model_pricing_table() -> Dict[str, dict]:
|
||
"""返回模型定价表（供前端展示）。"""
|
||
return {
|
||
model: {
|
||
"input_per_1M": input_p,
|
||
"output_per_1M": output_p,
|
||
}
|
||
for model, (input_p, output_p) in MODEL_PRICING.items()
|
||
}
|