mirror of
https://github.com/danny-avila/LibreChat.git
synced 2026-03-04 23:30:19 +01:00
💰 feat: Add gpt-5.3 context window and pricing (#12049)
* 💰 feat: Add gpt-5.3 context window and pricing * 💰 feat: Add OpenAI cached input pricing and `gpt-5.2-pro` model - Add cached input pricing (write/read) for gpt-4o, gpt-4.1, gpt-5.x, o1, o3, o4-mini models with correct per-family discount tiers - Add gpt-5.2-pro pricing ($21/$168), context window, and max output - Pro models (gpt-5-pro, gpt-5.2-pro) correctly excluded from cache pricing as OpenAI does not support caching for these * 🔍 fix: Address review findings for OpenAI pricing - Add o1-preview to cacheTokenValues (50% discount, same as o1) - Fix comment to enumerate all models per discount tier - Add cache tests for dated variants (gpt-4o-2024-08-06, etc.) - Add gpt-5-mini/gpt-5-nano to 10% ratio invariant test - Replace forEach with for...of in new test code - Fix inconsistent test description phrasing - Add gpt-5.3-preview to context window tests
This commit is contained in:
Danny Avila
GitHub
parent
474001c140
commit
d3622844ad
4 changed files with 156 additions and 2 deletions
|
|
@ -150,9 +150,11 @@ const tokenValues = Object.assign(
|
|||
'gpt-5': { prompt: 1.25, completion: 10 },
|
||||
'gpt-5.1': { prompt: 1.25, completion: 10 },
|
||||
'gpt-5.2': { prompt: 1.75, completion: 14 },
|
||||
'gpt-5.3': { prompt: 1.75, completion: 14 },
|
||||
'gpt-5-nano': { prompt: 0.05, completion: 0.4 },
|
||||
'gpt-5-mini': { prompt: 0.25, completion: 2 },
|
||||
'gpt-5-pro': { prompt: 15, completion: 120 },
|
||||
'gpt-5.2-pro': { prompt: 21, completion: 168 },
|
||||
o1: { prompt: 15, completion: 60 },
|
||||
'o1-mini': { prompt: 1.1, completion: 4.4 },
|
||||
'o1-preview': { prompt: 15, completion: 60 },
|
||||
|
|
@ -316,6 +318,28 @@ const cacheTokenValues = {
|
|||
'claude-opus-4': { write: 18.75, read: 1.5 },
|
||||
'claude-opus-4-5': { write: 6.25, read: 0.5 },
|
||||
'claude-opus-4-6': { write: 6.25, read: 0.5 },
|
||||
// OpenAI models — cached input discount varies by family:
|
||||
// gpt-4o (incl. mini), o1 (incl. mini/preview): 50% off
|
||||
// gpt-4.1 (incl. mini/nano), o3 (incl. mini), o4-mini: 75% off
|
||||
// gpt-5.x (excl. pro variants): 90% off
|
||||
// gpt-5-pro, gpt-5.2-pro: no caching
|
||||
'gpt-4o': { write: 2.5, read: 1.25 },
|
||||
'gpt-4o-mini': { write: 0.15, read: 0.075 },
|
||||
'gpt-4.1': { write: 2, read: 0.5 },
|
||||
'gpt-4.1-mini': { write: 0.4, read: 0.1 },
|
||||
'gpt-4.1-nano': { write: 0.1, read: 0.025 },
|
||||
'gpt-5': { write: 1.25, read: 0.125 },
|
||||
'gpt-5.1': { write: 1.25, read: 0.125 },
|
||||
'gpt-5.2': { write: 1.75, read: 0.175 },
|
||||
'gpt-5.3': { write: 1.75, read: 0.175 },
|
||||
'gpt-5-mini': { write: 0.25, read: 0.025 },
|
||||
'gpt-5-nano': { write: 0.05, read: 0.005 },
|
||||
o1: { write: 15, read: 7.5 },
|
||||
'o1-mini': { write: 1.1, read: 0.55 },
|
||||
'o1-preview': { write: 15, read: 7.5 },
|
||||
o3: { write: 2, read: 0.5 },
|
||||
'o3-mini': { write: 1.1, read: 0.275 },
|
||||
'o4-mini': { write: 1.1, read: 0.275 },
|
||||
// DeepSeek models - cache hit: $0.028/1M, cache miss: $0.28/1M
|
||||
deepseek: { write: 0.28, read: 0.028 },
|
||||
'deepseek-chat': { write: 0.28, read: 0.028 },
|
||||
|
|
|
|||
|
|
@ -52,6 +52,13 @@ describe('getValueKey', () => {
|
|||
expect(getValueKey('openai/gpt-5.2')).toBe('gpt-5.2');
|
||||
});
|
||||
|
||||
it('should return "gpt-5.3" for model name containing "gpt-5.3"', () => {
|
||||
expect(getValueKey('gpt-5.3')).toBe('gpt-5.3');
|
||||
expect(getValueKey('gpt-5.3-chat-latest')).toBe('gpt-5.3');
|
||||
expect(getValueKey('gpt-5.3-codex')).toBe('gpt-5.3');
|
||||
expect(getValueKey('openai/gpt-5.3')).toBe('gpt-5.3');
|
||||
});
|
||||
|
||||
it('should return "gpt-3.5-turbo-1106" for model name containing "gpt-3.5-turbo-1106"', () => {
|
||||
expect(getValueKey('gpt-3.5-turbo-1106-some-other-info')).toBe('gpt-3.5-turbo-1106');
|
||||
expect(getValueKey('openai/gpt-3.5-turbo-1106')).toBe('gpt-3.5-turbo-1106');
|
||||
|
|
@ -138,6 +145,12 @@ describe('getValueKey', () => {
|
|||
expect(getValueKey('gpt-5-pro-preview')).toBe('gpt-5-pro');
|
||||
});
|
||||
|
||||
it('should return "gpt-5.2-pro" for model name containing "gpt-5.2-pro"', () => {
|
||||
expect(getValueKey('gpt-5.2-pro')).toBe('gpt-5.2-pro');
|
||||
expect(getValueKey('gpt-5.2-pro-2025-03-01')).toBe('gpt-5.2-pro');
|
||||
expect(getValueKey('openai/gpt-5.2-pro')).toBe('gpt-5.2-pro');
|
||||
});
|
||||
|
||||
it('should return "gpt-4o" for model type of "gpt-4o"', () => {
|
||||
expect(getValueKey('gpt-4o-2024-08-06')).toBe('gpt-4o');
|
||||
expect(getValueKey('gpt-4o-2024-08-06-0718')).toBe('gpt-4o');
|
||||
|
|
@ -336,6 +349,18 @@ describe('getMultiplier', () => {
|
|||
);
|
||||
});
|
||||
|
||||
it('should return the correct multiplier for gpt-5.2-pro', () => {
|
||||
expect(getMultiplier({ model: 'gpt-5.2-pro', tokenType: 'prompt' })).toBe(
|
||||
tokenValues['gpt-5.2-pro'].prompt,
|
||||
);
|
||||
expect(getMultiplier({ model: 'gpt-5.2-pro', tokenType: 'completion' })).toBe(
|
||||
tokenValues['gpt-5.2-pro'].completion,
|
||||
);
|
||||
expect(getMultiplier({ model: 'openai/gpt-5.2-pro', tokenType: 'prompt' })).toBe(
|
||||
tokenValues['gpt-5.2-pro'].prompt,
|
||||
);
|
||||
});
|
||||
|
||||
it('should return the correct multiplier for gpt-5.1', () => {
|
||||
expect(getMultiplier({ model: 'gpt-5.1', tokenType: 'prompt' })).toBe(
|
||||
tokenValues['gpt-5.1'].prompt,
|
||||
|
|
@ -360,6 +385,21 @@ describe('getMultiplier', () => {
|
|||
);
|
||||
});
|
||||
|
||||
it('should return the correct multiplier for gpt-5.3', () => {
|
||||
expect(getMultiplier({ model: 'gpt-5.3', tokenType: 'prompt' })).toBe(
|
||||
tokenValues['gpt-5.3'].prompt,
|
||||
);
|
||||
expect(getMultiplier({ model: 'gpt-5.3', tokenType: 'completion' })).toBe(
|
||||
tokenValues['gpt-5.3'].completion,
|
||||
);
|
||||
expect(getMultiplier({ model: 'gpt-5.3-codex', tokenType: 'prompt' })).toBe(
|
||||
tokenValues['gpt-5.3'].prompt,
|
||||
);
|
||||
expect(getMultiplier({ model: 'openai/gpt-5.3', tokenType: 'completion' })).toBe(
|
||||
tokenValues['gpt-5.3'].completion,
|
||||
);
|
||||
});
|
||||
|
||||
it('should return the correct multiplier for gpt-4o', () => {
|
||||
const valueKey = getValueKey('gpt-4o-2024-08-06');
|
||||
expect(getMultiplier({ valueKey, tokenType: 'prompt' })).toBe(tokenValues['gpt-4o'].prompt);
|
||||
|
|
@ -1326,6 +1366,62 @@ describe('getCacheMultiplier', () => {
|
|||
).toBeNull();
|
||||
});
|
||||
|
||||
it('should return correct cache multipliers for OpenAI models', () => {
|
||||
const openaiCacheModels = [
|
||||
'gpt-4o',
|
||||
'gpt-4o-mini',
|
||||
'gpt-4.1',
|
||||
'gpt-4.1-mini',
|
||||
'gpt-4.1-nano',
|
||||
'gpt-5',
|
||||
'gpt-5.1',
|
||||
'gpt-5.2',
|
||||
'gpt-5.3',
|
||||
'gpt-5-mini',
|
||||
'gpt-5-nano',
|
||||
'o1',
|
||||
'o1-mini',
|
||||
'o1-preview',
|
||||
'o3',
|
||||
'o3-mini',
|
||||
'o4-mini',
|
||||
];
|
||||
|
||||
for (const model of openaiCacheModels) {
|
||||
expect(getCacheMultiplier({ model, cacheType: 'write' })).toBe(cacheTokenValues[model].write);
|
||||
expect(getCacheMultiplier({ model, cacheType: 'read' })).toBe(cacheTokenValues[model].read);
|
||||
}
|
||||
});
|
||||
|
||||
it('should return correct cache multipliers for OpenAI dated variants', () => {
|
||||
expect(getCacheMultiplier({ model: 'gpt-4o-2024-08-06', cacheType: 'read' })).toBe(
|
||||
cacheTokenValues['gpt-4o'].read,
|
||||
);
|
||||
expect(getCacheMultiplier({ model: 'gpt-4.1-2026-01-01', cacheType: 'read' })).toBe(
|
||||
cacheTokenValues['gpt-4.1'].read,
|
||||
);
|
||||
expect(getCacheMultiplier({ model: 'gpt-5.3-codex', cacheType: 'read' })).toBe(
|
||||
cacheTokenValues['gpt-5.3'].read,
|
||||
);
|
||||
expect(getCacheMultiplier({ model: 'openai/gpt-5.3', cacheType: 'write' })).toBe(
|
||||
cacheTokenValues['gpt-5.3'].write,
|
||||
);
|
||||
});
|
||||
|
||||
it('should return null for pro models that do not support caching', () => {
|
||||
expect(getCacheMultiplier({ model: 'gpt-5-pro', cacheType: 'read' })).toBeNull();
|
||||
expect(getCacheMultiplier({ model: 'gpt-5-pro', cacheType: 'write' })).toBeNull();
|
||||
expect(getCacheMultiplier({ model: 'gpt-5.2-pro', cacheType: 'read' })).toBeNull();
|
||||
expect(getCacheMultiplier({ model: 'gpt-5.2-pro', cacheType: 'write' })).toBeNull();
|
||||
});
|
||||
|
||||
it('should have consistent 10% cache read pricing for gpt-5.x models', () => {
|
||||
const gpt5CacheModels = ['gpt-5', 'gpt-5.1', 'gpt-5.2', 'gpt-5.3', 'gpt-5-mini', 'gpt-5-nano'];
|
||||
for (const model of gpt5CacheModels) {
|
||||
expect(cacheTokenValues[model].read).toBeCloseTo(cacheTokenValues[model].write * 0.1, 10);
|
||||
}
|
||||
});
|
||||
|
||||
it('should handle models with "bedrock/" prefix', () => {
|
||||
expect(
|
||||
getCacheMultiplier({
|
||||
|
|
|
|||
|
|
@ -200,6 +200,20 @@ describe('getModelMaxTokens', () => {
|
|||
);
|
||||
});
|
||||
|
||||
test('should return correct tokens for gpt-5.3 matches', () => {
|
||||
expect(getModelMaxTokens('gpt-5.3')).toBe(maxTokensMap[EModelEndpoint.openAI]['gpt-5.3']);
|
||||
expect(getModelMaxTokens('gpt-5.3-codex')).toBe(maxTokensMap[EModelEndpoint.openAI]['gpt-5.3']);
|
||||
expect(getModelMaxTokens('openai/gpt-5.3')).toBe(
|
||||
maxTokensMap[EModelEndpoint.openAI]['gpt-5.3'],
|
||||
);
|
||||
expect(getModelMaxTokens('gpt-5.3-2025-03-01')).toBe(
|
||||
maxTokensMap[EModelEndpoint.openAI]['gpt-5.3'],
|
||||
);
|
||||
expect(getModelMaxTokens('gpt-5.3-preview')).toBe(
|
||||
maxTokensMap[EModelEndpoint.openAI]['gpt-5.3'],
|
||||
);
|
||||
});
|
||||
|
||||
test('should return correct tokens for Anthropic models', () => {
|
||||
const models = [
|
||||
'claude-2.1',
|
||||
|
|
@ -492,7 +506,17 @@ describe('getModelMaxTokens', () => {
|
|||
|
||||
test('should return correct max output tokens for GPT-5 models', () => {
|
||||
const { getModelMaxOutputTokens } = require('@librechat/api');
|
||||
['gpt-5', 'gpt-5-mini', 'gpt-5-nano', 'gpt-5-pro'].forEach((model) => {
|
||||
const gpt5Models = [
|
||||
'gpt-5',
|
||||
'gpt-5.1',
|
||||
'gpt-5.2',
|
||||
'gpt-5.3',
|
||||
'gpt-5-mini',
|
||||
'gpt-5-nano',
|
||||
'gpt-5-pro',
|
||||
'gpt-5.2-pro',
|
||||
];
|
||||
for (const model of gpt5Models) {
|
||||
expect(getModelMaxOutputTokens(model)).toBe(maxOutputTokensMap[EModelEndpoint.openAI][model]);
|
||||
expect(getModelMaxOutputTokens(model, EModelEndpoint.openAI)).toBe(
|
||||
maxOutputTokensMap[EModelEndpoint.openAI][model],
|
||||
|
|
@ -500,7 +524,7 @@ describe('getModelMaxTokens', () => {
|
|||
expect(getModelMaxOutputTokens(model, EModelEndpoint.azureOpenAI)).toBe(
|
||||
maxOutputTokensMap[EModelEndpoint.azureOpenAI][model],
|
||||
);
|
||||
});
|
||||
}
|
||||
});
|
||||
|
||||
test('should return correct max output tokens for GPT-OSS models', () => {
|
||||
|
|
@ -612,6 +636,12 @@ describe('matchModelName', () => {
|
|||
expect(matchModelName('gpt-5-pro-2025-01-30-0130')).toBe('gpt-5-pro');
|
||||
});
|
||||
|
||||
it('should return the closest matching key for gpt-5.3 matches', () => {
|
||||
expect(matchModelName('openai/gpt-5.3')).toBe('gpt-5.3');
|
||||
expect(matchModelName('gpt-5.3-codex')).toBe('gpt-5.3');
|
||||
expect(matchModelName('gpt-5.3-2025-03-01')).toBe('gpt-5.3');
|
||||
});
|
||||
|
||||
// Tests for Google models
|
||||
it('should return the exact model name if it exists in maxTokensMap - Google models', () => {
|
||||
expect(matchModelName('text-bison-32k', EModelEndpoint.google)).toBe('text-bison-32k');
|
||||
|
|
|
|||
|
|
@ -51,9 +51,11 @@ const openAIModels = {
|
|||
'gpt-5': 400000,
|
||||
'gpt-5.1': 400000,
|
||||
'gpt-5.2': 400000,
|
||||
'gpt-5.3': 400000,
|
||||
'gpt-5-mini': 400000,
|
||||
'gpt-5-nano': 400000,
|
||||
'gpt-5-pro': 400000,
|
||||
'gpt-5.2-pro': 400000,
|
||||
'gpt-4o': 127500, // -500 from max
|
||||
'gpt-4o-mini': 127500, // -500 from max
|
||||
'gpt-4o-2024-05-13': 127500, // -500 from max
|
||||
|
|
@ -389,9 +391,11 @@ export const modelMaxOutputs = {
|
|||
'gpt-5': 128000,
|
||||
'gpt-5.1': 128000,
|
||||
'gpt-5.2': 128000,
|
||||
'gpt-5.3': 128000,
|
||||
'gpt-5-mini': 128000,
|
||||
'gpt-5-nano': 128000,
|
||||
'gpt-5-pro': 128000,
|
||||
'gpt-5.2-pro': 128000,
|
||||
'gpt-oss-20b': 131000,
|
||||
'gpt-oss-120b': 131000,
|
||||
system_default: 32000,
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue