Spaces:
Runtime error
Runtime error
File size: 4,284 Bytes
cd8bd0a | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 | /**
* Cost Estimator — Pre-flight cost estimation for LLM requests
*
* Estimates token-based costs before routing to a provider.
* Uses pricing data from the dashboard/database.
*
* @module shared/utils/costEstimator
*/
import { formatCost } from "./formatting";
export { formatCost };
/**
* Default pricing per 1M tokens (fallback when no pricing config exists).
* Values in USD.
*/
const DEFAULT_PRICING = {
"gpt-4o": { input: 2.5, output: 10.0 },
"gpt-4o-mini": { input: 0.15, output: 0.6 },
"gpt-4.1": { input: 2.0, output: 8.0 },
"gpt-4.1-mini": { input: 0.4, output: 1.6 },
"gpt-4.1-nano": { input: 0.1, output: 0.4 },
o3: { input: 2.0, output: 8.0 },
"o4-mini": { input: 1.1, output: 4.4 },
"claude-sonnet-4-5-20250514": { input: 3.0, output: 15.0 },
"claude-3-5-haiku-20241022": { input: 0.8, output: 4.0 },
"gemini-2.5-pro": { input: 1.25, output: 10.0 },
"gemini-2.5-flash": { input: 0.15, output: 0.6 },
};
/**
* Rough token estimation from text.
* Uses ~4 chars per token approximation (GPT-family average).
*
* @param {string} text
* @returns {number} Estimated token count
*/
export function estimateTokens(text) {
if (!text || typeof text !== "string") return 0;
return Math.ceil(text.length / 4);
}
/**
* Estimate input tokens from a chat completion request body.
*
* @param {Object} body - Request body
* @param {Array<{role: string, content: string|Array<{type: string, text?: string}>}>} [body.messages]
* @param {string} [body.system]
* @returns {number} Estimated input token count
*/
export function estimateInputTokens(body) {
if (!body) return 0;
let total = 0;
if (body.system) total += estimateTokens(body.system);
if (Array.isArray(body.messages)) {
for (const msg of body.messages) {
if (typeof msg.content === "string") {
total += estimateTokens(msg.content);
} else if (Array.isArray(msg.content)) {
for (const part of msg.content) {
if (part.type === "text" && typeof part.text === "string") {
total += estimateTokens(part.text);
}
}
}
// Add ~4 tokens overhead per message (role, separators)
total += 4;
}
}
return total;
}
/**
* Estimate the cost of a request given a model.
*
* @param {Object} params
* @param {string} params.model - Model identifier
* @param {number} params.inputTokens - Estimated input tokens
* @param {number} [params.maxOutputTokens=1000] - Max output tokens
* @param {Object} [params.pricingOverrides] - Custom pricing { input, output } per 1M tokens
* @returns {{ inputCost: number, outputCost: number, totalCost: number, model: string, inputTokens: number, outputTokens: number }}
*/
export function estimateCost({ model, inputTokens, maxOutputTokens = 1000, pricingOverrides }) {
// Find matching pricing (exact match or prefix match)
let pricing = pricingOverrides;
if (!pricing) {
const key = Object.keys(DEFAULT_PRICING).find((k) => model === k || model.startsWith(k));
pricing = key ? DEFAULT_PRICING[key] : { input: 1.0, output: 3.0 }; // conservative fallback
}
const inputCost = (inputTokens / 1_000_000) * pricing.input;
const outputCost = (maxOutputTokens / 1_000_000) * pricing.output;
const totalCost = inputCost + outputCost;
return {
model,
inputTokens,
outputTokens: maxOutputTokens,
inputCost: Math.round(inputCost * 1_000_000) / 1_000_000,
outputCost: Math.round(outputCost * 1_000_000) / 1_000_000,
totalCost: Math.round(totalCost * 1_000_000) / 1_000_000,
};
}
/**
* Quick pre-flight estimate: given a request body and model, return estimated cost.
*
* @param {Object} body - Chat completion request body
* @param {string} model - Target model
* @param {Object} [pricingOverrides] - Optional pricing overrides
* @returns {{ inputCost: number, outputCost: number, totalCost: number, formatted: string }}
*/
export function preflightEstimate(body, model, pricingOverrides) {
const inputTokens = estimateInputTokens(body);
const maxOutput = body.max_tokens || body.maxOutputTokens || 1000;
const result = estimateCost({ model, inputTokens, maxOutputTokens: maxOutput, pricingOverrides });
return {
...result,
formatted: formatCost(result.totalCost),
};
}
|