mirror of
https://github.com/enricoros/big-AGI.git
synced 2026-05-11 14:10:15 -07:00
Enrico Ros
1223 lines
44 KiB
TypeScript
1223 lines
44 KiB
TypeScript
import type { OpenAIWire_API_Models_List } from '~/modules/aix/server/dispatch/wiretypes/openai.wiretypes';
|
|
|
|
import { DModelInterfaceV1, LLM_IF_HOTFIX_NoTemperature, LLM_IF_HOTFIX_StripImages, LLM_IF_OAI_Chat, LLM_IF_OAI_Fn, LLM_IF_OAI_Json, LLM_IF_OAI_PromptCaching, LLM_IF_OAI_Reasoning, LLM_IF_OAI_Responses, LLM_IF_OAI_Vision, LLM_IF_Outputs_Audio, LLM_IF_Tools_WebSearch } from '~/common/stores/llms/llms.types';
|
|
import { Release } from '~/common/app.release';
|
|
|
|
import type { ModelDescriptionSchema } from '../../llm.server.types';
|
|
import { createVariantInjector, ModelVariantMap } from '../../llm.server.variants';
|
|
import { fromManualMapping, KnownModel, llmDevCheckModels_DEV, ManualMappings } from '../../models.mappings';
|
|
|
|
|
|
// OpenAI Model Variants
|
|
export const hardcodedOpenAIVariants: ModelVariantMap = {
|
|
|
|
// GPT-5.2 with reasoning disabled (non-thinking) - supports temperature control
|
|
// Per https://platform.openai.com/docs/guides/latest-model#gpt-5-2-parameter-compatibility
|
|
// temperature, top_p, logprobs are only supported when reasoning_effort=none
|
|
'gpt-5.2-2025-12-11': {
|
|
idVariant: '::thinking-none',
|
|
label: 'GPT-5.2 (No-thinking)',
|
|
hidden: true, // hidden by default as redundant, user can unhide in settings
|
|
description: 'Supports temperature control for creative applications. GPT-5.2 with reasoning disabled (reasoning_effort=none).',
|
|
interfaces: [LLM_IF_OAI_Responses, LLM_IF_OAI_Chat, LLM_IF_OAI_Vision, LLM_IF_OAI_Fn, LLM_IF_OAI_Json, LLM_IF_OAI_PromptCaching, LLM_IF_Tools_WebSearch], // NO LLM_IF_OAI_Reasoning, NO LLM_IF_HOTFIX_NoTemperature
|
|
parameterSpecs: [
|
|
{ paramId: 'llmVndOaiReasoningEffort52', initialValue: 'none', hidden: true }, // factory 'none', not changeable
|
|
{ paramId: 'llmVndOaiWebSearchContext' },
|
|
{ paramId: 'llmVndOaiVerbosity' },
|
|
{ paramId: 'llmVndOaiImageGeneration' },
|
|
{ paramId: 'llmVndOaiCodeInterpreter' },
|
|
{ paramId: 'llmForceNoStream' },
|
|
],
|
|
},
|
|
|
|
// GPT-5 with web search enabled by default
|
|
// 'gpt-5-2025-08-07': {
|
|
// idVariant: 'search',
|
|
// label: 'GPT-5 + Search',
|
|
// description: 'GPT-5 with web search enabled by default for up-to-date information and research.',
|
|
// parameterSpecs: [
|
|
// // customize this param
|
|
// { paramId: 'llmVndOaiWebSearchContext', initialValue: 'medium', hidden: true }, // Search enabled by default
|
|
// // copy other params
|
|
// { paramId: 'llmVndOaiReasoningEffort4' },
|
|
// { paramId: 'llmVndOaiRestoreMarkdown' },
|
|
// { paramId: 'llmVndOaiVerbosity' },
|
|
// { paramId: 'llmVndOaiImageGeneration' },
|
|
// ],
|
|
// benchmark: { cbaElo: 1442 + 1 }, // +1 from base GPT-5
|
|
// },
|
|
|
|
} as const;
|
|
|
|
|
|
// configuration
|
|
const DEV_DEBUG_OPENAI_MODELS = Release.IsNodeDevBuild; // not in staging to reduce noise
|
|
|
|
|
|
// per-family interfaces
|
|
const IFS_GPT_AUDIO: DModelInterfaceV1[] = [LLM_IF_OAI_Chat, LLM_IF_Outputs_Audio] as const;
|
|
const IFS_CHAT_MIN: DModelInterfaceV1[] = [LLM_IF_OAI_Chat, LLM_IF_OAI_Vision, LLM_IF_OAI_Fn, LLM_IF_OAI_Json] as const;
|
|
const IFS_CHAT_CACHE: DModelInterfaceV1[] = [LLM_IF_OAI_Chat, LLM_IF_OAI_Vision, LLM_IF_OAI_Fn, LLM_IF_OAI_Json, LLM_IF_OAI_PromptCaching] as const;
|
|
const IFS_CHAT_CACHE_REASON: DModelInterfaceV1[] = [LLM_IF_OAI_Chat, LLM_IF_OAI_Vision, LLM_IF_OAI_Fn, LLM_IF_OAI_Json, LLM_IF_OAI_PromptCaching, LLM_IF_OAI_Reasoning] as const;
|
|
|
|
// per-type parameter specs
|
|
const PS_DEEP_RESEARCH = [{ paramId: 'llmVndOaiWebSearchContext' as const, initialValue: 'medium', hidden: true } as const];
|
|
|
|
|
|
// [OpenAI] Known Chat Models
|
|
// https://platform.openai.com/docs/models
|
|
// https://platform.openai.com/docs/pricing
|
|
// NOTES:
|
|
// - "Structured Outputs" is LLM_IF_OAI_Json
|
|
export const _knownOpenAIChatModels: ManualMappings = [
|
|
|
|
/// GPT-5.2 series - Released December 11, 2025
|
|
|
|
// GPT-5.2
|
|
{
|
|
idPrefix: 'gpt-5.2-2025-12-11',
|
|
label: 'GPT-5.2 (2025-12-11)',
|
|
description: 'Most capable model for professional work and long-running agents. Improvements in general intelligence, long-context, agentic tool-calling, and vision.',
|
|
contextWindow: 400000,
|
|
maxCompletionTokens: 128000,
|
|
trainingDataCutoff: 'Aug 2025',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON, LLM_IF_Tools_WebSearch, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [
|
|
{ paramId: 'llmVndOaiReasoningEffort52', initialValue: 'medium' /* our decision: set to medium to have thinking - clones can set to 'none' to have temperature */ },
|
|
{ paramId: 'llmVndOaiWebSearchContext' },
|
|
{ paramId: 'llmVndOaiVerbosity' },
|
|
{ paramId: 'llmVndOaiImageGeneration' },
|
|
{ paramId: 'llmVndOaiCodeInterpreter' },
|
|
{ paramId: 'llmForceNoStream' },
|
|
],
|
|
chatPrice: { input: 1.75, cache: { cType: 'oai-ac', read: 0.175 }, output: 14 },
|
|
// benchmark: TBD
|
|
},
|
|
{
|
|
idPrefix: 'gpt-5.2',
|
|
label: 'GPT-5.2',
|
|
symLink: 'gpt-5.2-2025-12-11',
|
|
},
|
|
|
|
// GPT-5.2 Codex
|
|
{
|
|
idPrefix: 'gpt-5.2-codex',
|
|
label: 'GPT-5.2 Codex',
|
|
description: 'GPT-5.2 optimized for long-horizon, agentic coding tasks in Codex or similar environments. Supports low, medium, high, and xhigh reasoning effort settings.',
|
|
contextWindow: 400000,
|
|
maxCompletionTokens: 128000,
|
|
trainingDataCutoff: 'Aug 2025',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [
|
|
{ paramId: 'llmForceNoStream' },
|
|
{ paramId: 'llmVndOaiReasoningEffort52' },
|
|
{ paramId: 'llmVndOaiWebSearchContext' },
|
|
],
|
|
chatPrice: { input: 1.75, cache: { cType: 'oai-ac', read: 0.175 }, output: 14 },
|
|
// benchmark: TBD
|
|
},
|
|
|
|
// GPT-5.2 Chat Latest
|
|
{
|
|
idPrefix: 'gpt-5.2-chat-latest',
|
|
label: 'GPT-5.2 Instant',
|
|
description: 'GPT-5.2 model powering ChatGPT. Fast, capable for everyday work with clear improvements in info-seeking, how-tos, technical writing.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
trainingDataCutoff: 'Aug 2025',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE, LLM_IF_Tools_WebSearch, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [
|
|
{ paramId: 'llmVndOaiWebSearchContext' },
|
|
// { paramId: 'llmVndOaiVerbosity' }, // 2026-01-20: still unsupported
|
|
{ paramId: 'llmVndOaiImageGeneration' },
|
|
{ paramId: 'llmVndOaiCodeInterpreter' },
|
|
],
|
|
chatPrice: { input: 1.75, cache: { cType: 'oai-ac', read: 0.175 }, output: 14 },
|
|
// benchmark: TBD
|
|
},
|
|
|
|
// GPT-5.2 Pro
|
|
{
|
|
idPrefix: 'gpt-5.2-pro-2025-12-11',
|
|
label: 'GPT-5.2 Pro (2025-12-11)',
|
|
description: 'Smartest and most trustworthy option for difficult questions. Uses more compute for harder thinking on complex domains like programming.',
|
|
contextWindow: 400000,
|
|
maxCompletionTokens: 272000,
|
|
trainingDataCutoff: 'Aug 2025',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_MIN, LLM_IF_OAI_Reasoning, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [
|
|
{ paramId: 'llmVndOaiReasoningEffort52Pro' },
|
|
{ paramId: 'llmVndOaiWebSearchContext' },
|
|
{ paramId: 'llmForceNoStream' },
|
|
],
|
|
chatPrice: { input: 21, output: 168 },
|
|
// benchmark: TBD
|
|
},
|
|
{
|
|
idPrefix: 'gpt-5.2-pro',
|
|
label: 'GPT-5.2 Pro',
|
|
symLink: 'gpt-5.2-pro-2025-12-11',
|
|
},
|
|
|
|
|
|
/// GPT-5.1 series - Released November 13, 2025
|
|
|
|
// GPT-5.1
|
|
{
|
|
idPrefix: 'gpt-5.1-2025-11-13',
|
|
label: 'GPT-5.1 (2025-11-13)',
|
|
description: 'The best model for coding and agentic tasks with configurable reasoning effort.',
|
|
contextWindow: 400000,
|
|
maxCompletionTokens: 128000,
|
|
// trainingDataCutoff: 'Sep 30, 2024',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON, LLM_IF_Tools_WebSearch, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [
|
|
{ paramId: 'llmVndOaiReasoningEffort4' }, { paramId: 'llmVndOaiWebSearchContext' },
|
|
{ paramId: 'llmVndOaiVerbosity' },
|
|
{ paramId: 'llmVndOaiImageGeneration' },
|
|
{ paramId: 'llmVndOaiCodeInterpreter' },
|
|
{ paramId: 'llmForceNoStream' },
|
|
],
|
|
chatPrice: { input: 1.25, cache: { cType: 'oai-ac', read: 0.125 }, output: 10 },
|
|
// benchmark: TBD
|
|
},
|
|
{
|
|
idPrefix: 'gpt-5.1',
|
|
label: 'GPT-5.1',
|
|
symLink: 'gpt-5.1-2025-11-13',
|
|
},
|
|
|
|
// GPT-5.1 Chat Latest
|
|
{
|
|
idPrefix: 'gpt-5.1-chat-latest',
|
|
label: 'GPT-5.1 Instant',
|
|
description: 'GPT-5.1 Instant with adaptive reasoning. More conversational with improved instruction following.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
// trainingDataCutoff: 'Oct 31, 2024',
|
|
// interfaces: [LLM_IF_OAI_Responses, LLM_IF_OAI_Chat, LLM_IF_OAI_Vision, LLM_IF_OAI_PromptCaching], // no function calling or reasoning
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON, LLM_IF_Tools_WebSearch, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [
|
|
{ paramId: 'llmVndOaiWebSearchContext' },
|
|
// { paramId: 'llmVndOaiVerbosity' }, // 2026-01-20: still unsupported
|
|
{ paramId: 'llmVndOaiImageGeneration' },
|
|
{ paramId: 'llmVndOaiCodeInterpreter' },
|
|
],
|
|
chatPrice: { input: 1.25, cache: { cType: 'oai-ac', read: 0.125 }, output: 10 }, // TODO: Update with official pricing when available - this is `gpt-5-chat-latest` pricing
|
|
// benchmark: TBD
|
|
},
|
|
|
|
// GPT-5.1 Codex Max
|
|
{
|
|
idPrefix: 'gpt-5.1-codex-max',
|
|
label: 'GPT-5.1 Codex Max',
|
|
description: 'Our most intelligent coding model optimized for long-horizon, agentic coding tasks.',
|
|
contextWindow: 400000,
|
|
maxCompletionTokens: 128000,
|
|
trainingDataCutoff: 'Sep 30, 2024',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [
|
|
{ paramId: 'llmVndOaiReasoningEffort4' },
|
|
{ paramId: 'llmForceNoStream' },
|
|
],
|
|
chatPrice: { input: 1.25, cache: { cType: 'oai-ac', read: 0.125 }, output: 10 },
|
|
// benchmark: TBD
|
|
},
|
|
// GPT-5.1 Codex
|
|
{
|
|
idPrefix: 'gpt-5.1-codex',
|
|
label: 'GPT-5.1 Codex',
|
|
description: 'A version of GPT-5.1 optimized for agentic coding tasks in Codex or similar environments.',
|
|
contextWindow: 400000,
|
|
maxCompletionTokens: 128000,
|
|
trainingDataCutoff: 'Sep 30, 2024',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [
|
|
{ paramId: 'llmVndOaiReasoningEffort' }, // low, medium, high (no minimal)
|
|
{ paramId: 'llmForceNoStream' },
|
|
],
|
|
chatPrice: { input: 1.25, cache: { cType: 'oai-ac', read: 0.125 }, output: 10 },
|
|
// benchmark: TBD
|
|
},
|
|
// GPT-5.1 Codex Mini
|
|
{
|
|
idPrefix: 'gpt-5.1-codex-mini',
|
|
label: 'GPT-5.1 Codex Mini',
|
|
description: 'Smaller, faster version of GPT-5.1 Codex for efficient coding tasks.',
|
|
contextWindow: 400000,
|
|
maxCompletionTokens: 128000,
|
|
trainingDataCutoff: 'Sep 30, 2024',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [
|
|
{ paramId: 'llmVndOaiReasoningEffort' }, // low, medium, high (no minimal)
|
|
{ paramId: 'llmForceNoStream' },
|
|
],
|
|
chatPrice: { input: 0.25, cache: { cType: 'oai-ac', read: 0.025 }, output: 2 },
|
|
// benchmark: TBD
|
|
},
|
|
|
|
|
|
|
|
/// GPT-5 series - Released August 7, 2025
|
|
|
|
// GPT-5
|
|
{
|
|
idPrefix: 'gpt-5-2025-08-07',
|
|
label: 'GPT-5 (2025-08-07)',
|
|
description: 'The best model for coding and agentic tasks across domains.',
|
|
contextWindow: 400000,
|
|
maxCompletionTokens: 128000,
|
|
trainingDataCutoff: 'Sep 30, 2024',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON, LLM_IF_Tools_WebSearch, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [
|
|
{ paramId: 'llmVndOaiReasoningEffort4' }, { paramId: 'llmVndOaiWebSearchContext' },
|
|
{ paramId: 'llmVndOaiVerbosity' }, // gpt-5-class nets have verbosity control
|
|
{ paramId: 'llmVndOaiImageGeneration' }, // image generation capability
|
|
{ paramId: 'llmVndOaiCodeInterpreter' }, // code execution in sandboxed container
|
|
{ paramId: 'llmForceNoStream' }, // non-streaming option for unverified organizations
|
|
],
|
|
chatPrice: { input: 1.25, cache: { cType: 'oai-ac', read: 0.125 }, output: 10 },
|
|
benchmark: { cbaElo: 1442 }, // gpt-5-high
|
|
},
|
|
{
|
|
idPrefix: 'gpt-5',
|
|
label: 'GPT-5',
|
|
symLink: 'gpt-5-2025-08-07',
|
|
},
|
|
|
|
// GPT-5 Pro
|
|
{
|
|
idPrefix: 'gpt-5-pro-2025-10-06',
|
|
label: 'GPT-5 Pro (2025-10-06)',
|
|
description: 'Version of GPT-5 that uses more compute to produce smarter and more precise responses. Designed for tough problems.',
|
|
contextWindow: 400000,
|
|
maxCompletionTokens: 272000,
|
|
trainingDataCutoff: 'Sep 30, 2024',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_MIN, LLM_IF_OAI_Reasoning, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [{ paramId: 'llmForceNoStream' }], // reasoning effort is fixed at 'high'
|
|
chatPrice: { input: 15, output: 120 },
|
|
// benchmark: has not been measured yet
|
|
},
|
|
{
|
|
idPrefix: 'gpt-5-pro',
|
|
label: 'GPT-5 Pro',
|
|
symLink: 'gpt-5-pro-2025-10-06',
|
|
},
|
|
|
|
// GPT-5 Chat Latest
|
|
{
|
|
idPrefix: 'gpt-5-chat-latest',
|
|
label: 'GPT-5 ChatGPT (Non-Thinking)',
|
|
description: 'GPT-5 model used in ChatGPT. Points to the GPT-5 snapshot currently used in ChatGPT.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
trainingDataCutoff: 'Sep 29, 2024',
|
|
interfaces: [LLM_IF_OAI_Responses, LLM_IF_OAI_Chat, LLM_IF_OAI_Vision, LLM_IF_OAI_PromptCaching], // no function calling or reasoning
|
|
chatPrice: { input: 1.25, cache: { cType: 'oai-ac', read: 0.125 }, output: 10 },
|
|
benchmark: { cbaElo: 1430 }, // gpt-5-chat
|
|
},
|
|
|
|
// GPT-5 Codex
|
|
{
|
|
idPrefix: 'gpt-5-codex',
|
|
label: 'GPT-5 Codex',
|
|
description: 'A version of GPT-5 optimized for agentic coding in Codex.',
|
|
contextWindow: 400000,
|
|
maxCompletionTokens: 128000,
|
|
trainingDataCutoff: 'Sep 30, 2024',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON, LLM_IF_Tools_WebSearch, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [
|
|
{ paramId: 'llmVndOaiReasoningEffort' }, // works
|
|
{ paramId: 'llmVndOaiWebSearchContext' }, // works, although is not triggered often
|
|
// { paramId: 'llmVndOaiRestoreMarkdown', initialValue: false }, // since this is for code, let the prompt dictate markdown usage rather than us injecting
|
|
{ paramId: 'llmForceNoStream' },
|
|
],
|
|
chatPrice: { input: 1.25, cache: { cType: 'oai-ac', read: 0.125 }, output: 10 },
|
|
// benchmark: TBD
|
|
},
|
|
|
|
// GPT-5 Search API
|
|
{
|
|
idPrefix: 'gpt-5-search-api-2025-10-14',
|
|
label: 'GPT-5 Search API (2025-10-14)',
|
|
description: 'Updated web search model in Chat Completions API. 60% cheaper with domain filtering support.',
|
|
contextWindow: 400000,
|
|
maxCompletionTokens: 100000,
|
|
trainingDataCutoff: 'Sep 30, 2024',
|
|
interfaces: [...IFS_CHAT_MIN, LLM_IF_Tools_WebSearch],
|
|
parameterSpecs: [{ paramId: 'llmVndOaiWebSearchContext', initialValue: 'medium' }], // Search enabled by default
|
|
chatPrice: { input: 1.25, cache: { cType: 'oai-ac', read: 0.125 }, output: 10 },
|
|
// benchmark: TBD
|
|
},
|
|
{
|
|
idPrefix: 'gpt-5-search-api',
|
|
label: 'GPT-5 Search API',
|
|
symLink: 'gpt-5-search-api-2025-10-14',
|
|
},
|
|
|
|
// GPT-5 mini
|
|
{
|
|
idPrefix: 'gpt-5-mini-2025-08-07',
|
|
label: 'GPT-5 Mini (2025-08-07)',
|
|
description: 'A faster, more cost-efficient version of GPT-5 for well-defined tasks.',
|
|
contextWindow: 400000,
|
|
maxCompletionTokens: 128000,
|
|
trainingDataCutoff: 'May 30, 2024',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON, LLM_IF_Tools_WebSearch, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [{ paramId: 'llmVndOaiReasoningEffort4' }, { paramId: 'llmVndOaiWebSearchContext' }, { paramId: 'llmVndOaiVerbosity' }, { paramId: 'llmVndOaiImageGeneration' }, { paramId: 'llmForceNoStream' }],
|
|
chatPrice: { input: 0.25, cache: { cType: 'oai-ac', read: 0.025 }, output: 2 },
|
|
benchmark: { cbaElo: 1388 }, // gpt-5-mini-high
|
|
},
|
|
{
|
|
idPrefix: 'gpt-5-mini',
|
|
label: 'GPT-5 Mini',
|
|
symLink: 'gpt-5-mini-2025-08-07',
|
|
},
|
|
|
|
// GPT-5 nano
|
|
{
|
|
idPrefix: 'gpt-5-nano-2025-08-07',
|
|
label: 'GPT-5 Nano (2025-08-07)',
|
|
description: 'Fastest, most cost-efficient version of GPT-5 for summarization and classification tasks.',
|
|
contextWindow: 400000,
|
|
maxCompletionTokens: 128000,
|
|
trainingDataCutoff: 'May 30, 2024',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [{ paramId: 'llmVndOaiReasoningEffort4' }, { paramId: 'llmVndOaiVerbosity' }, { paramId: 'llmVndOaiImageGeneration' }],
|
|
chatPrice: { input: 0.05, cache: { cType: 'oai-ac', read: 0.005 }, output: 0.4 },
|
|
benchmark: { cbaElo: 1344 }, // gpt-5-nano-high
|
|
},
|
|
{
|
|
idPrefix: 'gpt-5-nano',
|
|
label: 'GPT-5 Nano',
|
|
symLink: 'gpt-5-nano-2025-08-07',
|
|
},
|
|
|
|
/// OSB-120b - Speculative support for new model appearing in API
|
|
{
|
|
idPrefix: 'osb-120b',
|
|
label: 'OSB-120B',
|
|
description: 'Speculative support for osb-120b model. Uses Responses API.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 32768,
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [
|
|
{ paramId: 'llmVndOaiReasoningEffort' },
|
|
{ paramId: 'llmForceNoStream' },
|
|
],
|
|
// chatPrice: TBD - unknown pricing
|
|
},
|
|
|
|
|
|
/// [OpenAI, 2025-03-11] NEW `v1/responses` API MODELS - UNSUPPORTED YET
|
|
|
|
// Computer Use Preview - INTERNAL MODEL FOR AGENTS - UNSUPPORTED YET
|
|
{
|
|
hidden: true, // UNSUPPORTED YET
|
|
idPrefix: 'computer-use-preview-2025-03-11',
|
|
label: 'Computer Use Preview (2025-03-11)',
|
|
description: 'Specialized model for computer use tool. Optimized for computer interaction capabilities.',
|
|
contextWindow: 8192,
|
|
maxCompletionTokens: 1024,
|
|
trainingDataCutoff: 'Sep 30, 2023',
|
|
interfaces: [LLM_IF_OAI_Responses, LLM_IF_OAI_Chat, LLM_IF_OAI_Fn, LLM_IF_HOTFIX_NoTemperature],
|
|
chatPrice: { input: 3, output: 12 },
|
|
isPreview: true,
|
|
},
|
|
{
|
|
idPrefix: 'computer-use-preview',
|
|
label: 'Computer Use Preview',
|
|
symLink: 'computer-use-preview-2025-03-11',
|
|
},
|
|
{
|
|
hidden: true, // yield to more capable
|
|
idPrefix: 'codex-mini-latest',
|
|
label: 'Codex Mini Latest',
|
|
description: 'Fast reasoning model optimized for the Codex CLI. A fine-tuned version of o4-mini for low-latency code Q&A and editing.',
|
|
contextWindow: 200000,
|
|
maxCompletionTokens: 100000,
|
|
trainingDataCutoff: 'May 31, 2024',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [{ paramId: 'llmVndOaiReasoningEffort' }],
|
|
chatPrice: { input: 1.5, cache: { cType: 'oai-ac', read: 0.375 }, output: 6 },
|
|
isLegacy: true, // Deprecated January 16, 2026.
|
|
},
|
|
|
|
|
|
/// Reasoning models - o-series
|
|
|
|
// o4-mini-deep-research - (v1/responses API)
|
|
{
|
|
idPrefix: 'o4-mini-deep-research-2025-06-26',
|
|
label: 'o4 Mini Deep Research (2025-06-26)',
|
|
description: 'Faster, more affordable deep research model for complex, multi-step research tasks.',
|
|
contextWindow: 200000,
|
|
maxCompletionTokens: 100000,
|
|
trainingDataCutoff: 'May 31, 2024',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON],
|
|
parameterSpecs: PS_DEEP_RESEARCH,
|
|
chatPrice: { input: 2, cache: { cType: 'oai-ac', read: 0.5 }, output: 8 },
|
|
},
|
|
{
|
|
idPrefix: 'o4-mini-deep-research',
|
|
label: 'o4 Mini Deep Research',
|
|
symLink: 'o4-mini-deep-research-2025-06-26',
|
|
},
|
|
|
|
/// o4-mini
|
|
{
|
|
idPrefix: 'o4-mini-2025-04-16',
|
|
label: 'o4 Mini (2025-04-16)',
|
|
description: 'Latest o4-mini model. Optimized for fast, effective reasoning with exceptionally efficient performance in coding and visual tasks.',
|
|
contextWindow: 200000,
|
|
maxCompletionTokens: 100000,
|
|
trainingDataCutoff: 'May 31, 2024',
|
|
interfaces: IFS_CHAT_CACHE_REASON,
|
|
parameterSpecs: [{ paramId: 'llmVndOaiReasoningEffort' }],
|
|
chatPrice: { input: 1.1, cache: { cType: 'oai-ac', read: 0.275 }, output: 4.4 },
|
|
benchmark: { cbaElo: 1393 }, // o4-mini-2025-04-16
|
|
},
|
|
{
|
|
idPrefix: 'o4-mini',
|
|
label: 'o4 Mini',
|
|
symLink: 'o4-mini-2025-04-16',
|
|
},
|
|
|
|
// o3-deep-research - (v1/responses API)
|
|
{
|
|
idPrefix: 'o3-deep-research-2025-06-26',
|
|
label: 'o3 Deep Research (2025-06-26)',
|
|
description: 'Our most powerful deep research model for complex, multi-step research tasks.',
|
|
contextWindow: 200000,
|
|
maxCompletionTokens: 100000,
|
|
trainingDataCutoff: 'May 31, 2024',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_CACHE_REASON],
|
|
parameterSpecs: PS_DEEP_RESEARCH,
|
|
chatPrice: { input: 10, cache: { cType: 'oai-ac', read: 2.5 }, output: 40 },
|
|
},
|
|
{
|
|
idPrefix: 'o3-deep-research',
|
|
label: 'o3 Deep Research',
|
|
symLink: 'o3-deep-research-2025-06-26',
|
|
},
|
|
|
|
// o3-pro - (v1/responses API)
|
|
{
|
|
idPrefix: 'o3-pro-2025-06-10',
|
|
label: 'o3 Pro (2025-06-10)',
|
|
description: 'Version of o3 with more compute for better responses. Provides consistently better answers for complex tasks.',
|
|
contextWindow: 200000,
|
|
maxCompletionTokens: 100000,
|
|
trainingDataCutoff: 'May 31, 2024',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_MIN, LLM_IF_OAI_Reasoning, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [{ paramId: 'llmVndOaiReasoningEffort' }, { paramId: 'llmForceNoStream' }],
|
|
chatPrice: { input: 20, output: 80 },
|
|
// benchmark: has not been measured yet
|
|
},
|
|
{
|
|
idPrefix: 'o3-pro',
|
|
label: 'o3 Pro',
|
|
symLink: 'o3-pro-2025-06-10',
|
|
},
|
|
|
|
/// o3
|
|
{
|
|
idPrefix: 'o3-2025-04-16',
|
|
label: 'o3 (2025-04-16)',
|
|
description: 'A well-rounded and powerful model across domains. Sets a new standard for math, science, coding, and visual reasoning tasks.',
|
|
contextWindow: 200000,
|
|
maxCompletionTokens: 100000,
|
|
trainingDataCutoff: 'May 31, 2024',
|
|
interfaces: IFS_CHAT_CACHE_REASON,
|
|
parameterSpecs: [{ paramId: 'llmVndOaiReasoningEffort' }, { paramId: 'llmForceNoStream' }],
|
|
chatPrice: { input: 2, cache: { cType: 'oai-ac', read: 0.5 }, output: 8 },
|
|
benchmark: { cbaElo: 1444 }, // o3-2025-04-16
|
|
},
|
|
{
|
|
idPrefix: 'o3',
|
|
label: 'o3',
|
|
symLink: 'o3-2025-04-16',
|
|
},
|
|
|
|
// o3-mini
|
|
{
|
|
idPrefix: 'o3-mini-2025-01-31',
|
|
label: 'o3 Mini (2025-01-31)',
|
|
description: 'Latest o3-mini model snapshot. High intelligence at the same cost and latency targets of o1-mini. Excels at science, math, and coding tasks.',
|
|
contextWindow: 200000,
|
|
maxCompletionTokens: 100000,
|
|
trainingDataCutoff: 'Oct 2023',
|
|
interfaces: [LLM_IF_OAI_Chat, LLM_IF_OAI_Fn, LLM_IF_OAI_Json, LLM_IF_OAI_PromptCaching, LLM_IF_OAI_Reasoning, LLM_IF_HOTFIX_StripImages],
|
|
parameterSpecs: [{ paramId: 'llmVndOaiReasoningEffort' }],
|
|
chatPrice: { input: 1.1, cache: { cType: 'oai-ac', read: 0.55 }, output: 4.4 },
|
|
benchmark: { cbaElo: 1347 }, // o3-mini (not using o3-mini-high here, as it seems too inflated)
|
|
},
|
|
{
|
|
idPrefix: 'o3-mini',
|
|
label: 'o3 Mini',
|
|
symLink: 'o3-mini-2025-01-31',
|
|
},
|
|
|
|
// o1-pro - (v1/responses API) 💎💰
|
|
{
|
|
hidden: true,
|
|
idPrefix: 'o1-pro-2025-03-19',
|
|
label: 'o1 Pro (2025-03-19)',
|
|
description: 'A version of o1 with more compute for better responses. Provides consistently better answers for complex tasks.',
|
|
contextWindow: 200000,
|
|
maxCompletionTokens: 100000,
|
|
trainingDataCutoff: 'Sep 30, 2023',
|
|
interfaces: [LLM_IF_OAI_Responses, ...IFS_CHAT_MIN, LLM_IF_OAI_Reasoning, LLM_IF_HOTFIX_NoTemperature],
|
|
parameterSpecs: [{ paramId: 'llmVndOaiReasoningEffort' }, { paramId: 'llmForceNoStream' }],
|
|
chatPrice: { input: 150, output: 600 },
|
|
// benchmark: has not been measured yet by third parties
|
|
},
|
|
{
|
|
idPrefix: 'o1-pro',
|
|
label: 'o1 Pro',
|
|
symLink: 'o1-pro-2025-03-19',
|
|
},
|
|
|
|
// o1
|
|
{
|
|
idPrefix: 'o1-2024-12-17',
|
|
label: 'o1 (2024-12-17)',
|
|
description: 'Previous full o-series reasoning model.',
|
|
contextWindow: 200000,
|
|
maxCompletionTokens: 100000,
|
|
trainingDataCutoff: 'Oct 2023',
|
|
interfaces: IFS_CHAT_CACHE_REASON,
|
|
parameterSpecs: [{ paramId: 'llmVndOaiReasoningEffort' }, { paramId: 'llmVndOaiRestoreMarkdown' }],
|
|
chatPrice: { input: 15, cache: { cType: 'oai-ac', read: 7.5 }, output: 60 },
|
|
benchmark: { cbaElo: 1399 }, // o1-2024-12-17
|
|
},
|
|
{
|
|
idPrefix: 'o1',
|
|
label: 'o1',
|
|
symLink: 'o1-2024-12-17',
|
|
},
|
|
|
|
|
|
/// GPT-4.1 series
|
|
|
|
// GPT-4.1
|
|
{
|
|
idPrefix: 'gpt-4.1-2025-04-14',
|
|
label: 'GPT-4.1 (2025-04-14)',
|
|
description: 'Flagship GPT model for complex tasks. Major improvements on coding, instruction following, and long context with 1M token context window.',
|
|
contextWindow: 1047576,
|
|
maxCompletionTokens: 32768,
|
|
trainingDataCutoff: 'May 31, 2024',
|
|
interfaces: IFS_CHAT_CACHE,
|
|
chatPrice: { input: 2, cache: { cType: 'oai-ac', read: 0.5 }, output: 8 },
|
|
benchmark: { cbaElo: 1409 }, // gpt-4.1-2025-04-14
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4.1',
|
|
label: 'GPT-4.1',
|
|
symLink: 'gpt-4.1-2025-04-14',
|
|
},
|
|
|
|
// GPT-4.1 mini
|
|
{
|
|
idPrefix: 'gpt-4.1-mini-2025-04-14',
|
|
label: 'GPT-4.1 Mini (2025-04-14)',
|
|
description: 'Balanced for intelligence, speed, and cost. Matches or exceeds GPT-4o in intelligence while reducing latency by nearly half and cost by 83%.',
|
|
contextWindow: 1047576,
|
|
maxCompletionTokens: 32768,
|
|
trainingDataCutoff: 'May 31, 2024',
|
|
interfaces: IFS_CHAT_CACHE,
|
|
chatPrice: { input: 0.4, cache: { cType: 'oai-ac', read: 0.1 }, output: 1.6 },
|
|
benchmark: { cbaElo: 1377 }, // gpt-4.1-mini-2025-04-14
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4.1-mini',
|
|
label: 'GPT-4.1 Mini',
|
|
symLink: 'gpt-4.1-mini-2025-04-14',
|
|
},
|
|
|
|
// GPT-4.1 nano
|
|
{
|
|
idPrefix: 'gpt-4.1-nano-2025-04-14',
|
|
label: 'GPT-4.1 Nano (2025-04-14)',
|
|
description: 'Fastest, most cost-effective GPT 4.1 model. Delivers exceptional performance with low latency, ideal for tasks like classification or autocompletion.',
|
|
contextWindow: 1047576,
|
|
maxCompletionTokens: 32768,
|
|
trainingDataCutoff: 'May 31, 2024',
|
|
interfaces: IFS_CHAT_CACHE,
|
|
chatPrice: { input: 0.1, cache: { cType: 'oai-ac', read: 0.025 }, output: 0.4 },
|
|
benchmark: { cbaElo: 1320 }, // gpt-4.1-nano-2025-04-14
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4.1-nano',
|
|
label: 'GPT-4.1 Nano',
|
|
symLink: 'gpt-4.1-nano-2025-04-14',
|
|
},
|
|
|
|
|
|
/// GPT-Audio series - General availability audio models
|
|
|
|
// gpt-audio
|
|
{
|
|
idPrefix: 'gpt-audio-2025-08-28',
|
|
label: 'GPT Audio (2025-08-28)',
|
|
description: 'First generally available audio model. Accepts audio inputs and outputs, and can be used in the Chat Completions REST API.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
interfaces: IFS_GPT_AUDIO,
|
|
chatPrice: { input: 2.5, output: 10 },
|
|
// benchmark: TBD
|
|
},
|
|
{
|
|
idPrefix: 'gpt-audio',
|
|
label: 'GPT Audio',
|
|
symLink: 'gpt-audio-2025-08-28',
|
|
},
|
|
{
|
|
idPrefix: 'gpt-audio-mini-2025-12-15',
|
|
label: 'GPT Audio Mini (2025-12-15)',
|
|
description: 'Cost-efficient audio model. Accepts audio inputs and outputs via Chat Completions REST API.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
interfaces: IFS_GPT_AUDIO,
|
|
chatPrice: { input: 0.6, output: 2.4 },
|
|
},
|
|
{
|
|
idPrefix: 'gpt-audio-mini-2025-10-06',
|
|
label: 'GPT Audio Mini (2025-10-06)',
|
|
hidden: true, // previous version
|
|
description: 'Cost-efficient audio model. Accepts audio inputs and outputs via Chat Completions REST API.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
interfaces: IFS_GPT_AUDIO,
|
|
chatPrice: { input: 0.6, output: 2.4 },
|
|
},
|
|
{
|
|
idPrefix: 'gpt-audio-mini',
|
|
label: 'GPT Audio Mini',
|
|
symLink: 'gpt-audio-mini-2025-12-15',
|
|
},
|
|
|
|
|
|
/// GPT-Realtime series - REMOVED
|
|
|
|
/// GPT-4/4o series
|
|
|
|
// GPT-4o
|
|
{
|
|
idPrefix: 'gpt-4o-2024-11-20',
|
|
label: 'GPT-4o (2024-11-20)',
|
|
description: 'Snapshot of gpt-4o from November 20th, 2024.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
trainingDataCutoff: 'Oct 2023',
|
|
interfaces: IFS_CHAT_CACHE,
|
|
chatPrice: { input: 2.5, cache: { cType: 'oai-ac', read: 1.25 }, output: 10 },
|
|
benchmark: { cbaElo: 1265 + 1 }, // not reported; using gpt-4o-2024-08-06 + 1
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4o-2024-08-06',
|
|
label: 'GPT-4o (2024-08-06)',
|
|
hidden: true, // previous version
|
|
description: 'Snapshot that supports Structured Outputs. gpt-4o currently points to this version.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
trainingDataCutoff: 'Oct 2023',
|
|
interfaces: IFS_CHAT_CACHE,
|
|
chatPrice: { input: 2.5, cache: { cType: 'oai-ac', read: 1.25 }, output: 10 },
|
|
benchmark: { cbaElo: 1333 }, // GPT-4o (08/06)
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4o-2024-05-13',
|
|
label: 'GPT-4o (2024-05-13)',
|
|
hidden: true, // previous version
|
|
description: 'Original gpt-4o snapshot from May 13, 2024.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 4096,
|
|
trainingDataCutoff: 'Oct 2023',
|
|
interfaces: IFS_CHAT_MIN,
|
|
chatPrice: { input: 5, output: 15 },
|
|
benchmark: { cbaElo: 1344 }, // gpt-4o-2024-05-13
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4o',
|
|
label: 'GPT-4o',
|
|
symLink: 'gpt-4o-2024-08-06',
|
|
},
|
|
{
|
|
idPrefix: 'chatgpt-4o-latest',
|
|
label: 'ChatGPT-4o Latest',
|
|
description: 'The chatgpt-4o-latest model version continuously points to the version of GPT-4o used in ChatGPT, and is updated frequently.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
trainingDataCutoff: 'Oct 2023',
|
|
interfaces: [LLM_IF_OAI_Chat, LLM_IF_OAI_Vision, LLM_IF_OAI_Json], // does not support Tools
|
|
chatPrice: { input: 5, output: 15 },
|
|
benchmark: { cbaElo: 1441 }, // chatgpt-4o-latest-20250326
|
|
isLegacy: true, // Deprecated February 17, 2026.
|
|
},
|
|
|
|
// GPT-4o Search Preview: When using Chat Completions, the model always retrieves information from the web before responding to your query.
|
|
{
|
|
hidden: true, // old
|
|
idPrefix: 'gpt-4o-search-preview-2025-03-11',
|
|
label: 'GPT-4o Search Preview (2025-03-11)',
|
|
description: 'Latest snapshot of the GPT-4o model optimized for web search capabilities.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
trainingDataCutoff: 'Sep 30, 2023',
|
|
interfaces: [LLM_IF_OAI_Chat, LLM_IF_OAI_Json, LLM_IF_Tools_WebSearch, LLM_IF_HOTFIX_NoTemperature], // NOTE: 2025-03-15: confirmed on 'playground' that this model does not support images
|
|
parameterSpecs: [{ paramId: 'llmVndOaiWebSearchContext' }, { paramId: 'llmVndOaiWebSearchGeolocation' }],
|
|
chatPrice: { input: 2.5, output: 10 },
|
|
// benchmarks don't apply to search models
|
|
isPreview: true,
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4o-search-preview',
|
|
label: 'GPT-4o Search Preview',
|
|
symLink: 'gpt-4o-search-preview-2025-03-11',
|
|
},
|
|
|
|
// GPT-4o Audio Preview
|
|
{
|
|
hidden: true, // old
|
|
idPrefix: 'gpt-4o-audio-preview-2025-06-03',
|
|
label: 'GPT-4o Audio Preview (2025-06-03)',
|
|
description: 'Latest snapshot for the Audio API model.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
trainingDataCutoff: 'Oct 2023',
|
|
interfaces: IFS_GPT_AUDIO,
|
|
chatPrice: { input: 2.5, output: 10 /* AUDIO PRICING UNSUPPORTED 40/80 */ },
|
|
// benchmarks don't apply to audio models
|
|
isPreview: true,
|
|
},
|
|
{
|
|
hidden: true, // old
|
|
idPrefix: 'gpt-4o-audio-preview-2024-12-17',
|
|
label: 'GPT-4o Audio Preview (2024-12-17)',
|
|
description: 'Snapshot for the Audio API model.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
trainingDataCutoff: 'Oct 2023',
|
|
interfaces: IFS_GPT_AUDIO,
|
|
chatPrice: { input: 2.5, output: 10 /* AUDIO PRICING UNSUPPORTED 40/80 */ },
|
|
// benchmarks don't apply to audio models
|
|
isPreview: true,
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4o-audio-preview',
|
|
label: 'GPT-4o Audio Preview',
|
|
symLink: 'gpt-4o-audio-preview-2025-06-03',
|
|
},
|
|
|
|
// GPT-4o mini
|
|
{
|
|
idPrefix: 'gpt-4o-mini-2024-07-18',
|
|
label: 'GPT-4o Mini (2024-07-18)',
|
|
description: 'Affordable model for fast, lightweight tasks. GPT-4o Mini is cheaper and more capable than GPT-3.5 Turbo.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
trainingDataCutoff: 'Oct 2023',
|
|
interfaces: IFS_CHAT_CACHE,
|
|
chatPrice: { input: 0.15, cache: { cType: 'oai-ac', read: 0.075 }, output: 0.6 },
|
|
benchmark: { cbaElo: 1316 }, // GPT-4o-mini (07/18)
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4o-mini',
|
|
label: 'GPT-4o mini',
|
|
symLink: 'gpt-4o-mini-2024-07-18',
|
|
},
|
|
{
|
|
hidden: true, // UNSUPPORTED yet (audio output model)
|
|
idPrefix: 'gpt-4o-mini-audio-preview-2024-12-17',
|
|
label: 'GPT-4o Mini Audio Preview (2024-12-17)',
|
|
description: 'Snapshot for the Audio API model.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
trainingDataCutoff: 'Oct 2023',
|
|
interfaces: IFS_GPT_AUDIO,
|
|
chatPrice: { input: 0.15, output: 0.6 /* AUDIO PRICING UNSUPPORTED 10/20 */ },
|
|
// benchmarks don't apply to audio models
|
|
isPreview: true,
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4o-mini-audio-preview',
|
|
label: 'GPT-4o Mini Audio Preview',
|
|
symLink: 'gpt-4o-mini-audio-preview-2024-12-17',
|
|
},
|
|
// GPT-4o Mini Search Preview: When using Chat Completions, the model always retrieves information from the web before responding to your query.
|
|
{
|
|
hidden: true, // old
|
|
idPrefix: 'gpt-4o-mini-search-preview-2025-03-11',
|
|
label: 'GPT-4o Mini Search Preview (2025-03-11)',
|
|
description: 'Latest snapshot of the GPT-4o Mini model optimized for web search capabilities.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 16384,
|
|
trainingDataCutoff: 'Sep 30, 2023',
|
|
interfaces: [LLM_IF_OAI_Chat, LLM_IF_OAI_Json, LLM_IF_Tools_WebSearch, LLM_IF_HOTFIX_NoTemperature], // NOTE: this support function calling, but only its own, not a Custom Function
|
|
parameterSpecs: [{ paramId: 'llmVndOaiWebSearchContext' }, { paramId: 'llmVndOaiWebSearchGeolocation' }],
|
|
chatPrice: { input: 0.15, output: 0.6 },
|
|
// benchmarks don't apply to search models
|
|
isPreview: true,
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4o-mini-search-preview',
|
|
label: 'GPT-4o Mini Search Preview',
|
|
symLink: 'gpt-4o-mini-search-preview-2025-03-11',
|
|
},
|
|
|
|
// GPT-4 Turbo
|
|
{
|
|
idPrefix: 'gpt-4-turbo-2024-04-09',
|
|
label: 'GPT-4 Turbo (2024-04-09)',
|
|
hidden: true, // OLD
|
|
description: 'GPT-4 Turbo with Vision model. Vision requests can now use JSON mode and function calling. gpt-4-turbo currently points to this version.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 4096,
|
|
trainingDataCutoff: 'Dec 2023',
|
|
interfaces: IFS_CHAT_MIN,
|
|
chatPrice: { input: 10, output: 30 },
|
|
benchmark: { cbaElo: 1324 }, // gpt-4-turbo-2024-04-09
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4-turbo',
|
|
label: 'GPT-4 Turbo',
|
|
symLink: 'gpt-4-turbo-2024-04-09',
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4-0125-preview',
|
|
label: 'GPT-4 Turbo (0125)',
|
|
hidden: true, // OLD
|
|
description: 'GPT-4 Turbo preview model intended to reduce cases of "laziness" where the model doesn\'t complete a task.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 4096,
|
|
trainingDataCutoff: 'Dec 2023',
|
|
interfaces: [LLM_IF_OAI_Chat, LLM_IF_OAI_Fn, LLM_IF_OAI_Json],
|
|
chatPrice: { input: 10, output: 30 },
|
|
benchmark: { cbaElo: 1315 }, // gpt-4-0125-preview
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4-1106-preview', // GPT-4 Turbo preview model
|
|
label: 'GPT-4 Turbo (1106)',
|
|
hidden: true, // OLD
|
|
description: 'GPT-4 Turbo preview model featuring improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 4096,
|
|
trainingDataCutoff: 'Apr 2023',
|
|
interfaces: [LLM_IF_OAI_Chat, LLM_IF_OAI_Fn, LLM_IF_OAI_Json],
|
|
chatPrice: { input: 10, output: 30 },
|
|
benchmark: { cbaElo: 1315 }, // gpt-4-1106-preview
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4-turbo-preview',
|
|
label: 'GPT-4 Turbo Preview',
|
|
symLink: 'gpt-4-0125-preview',
|
|
isLegacy: true,
|
|
},
|
|
|
|
// GPT4's
|
|
{
|
|
idPrefix: 'gpt-4-0613',
|
|
label: 'GPT-4 (0613)',
|
|
hidden: true, // OLD
|
|
description: 'Snapshot of gpt-4 from June 13th 2023 with improved function calling support. Data up to Sep 2021.',
|
|
contextWindow: 8192,
|
|
trainingDataCutoff: 'Sep 2021',
|
|
interfaces: [LLM_IF_OAI_Chat, LLM_IF_OAI_Fn],
|
|
chatPrice: { input: 30, output: 60 },
|
|
benchmark: { cbaElo: 1163 },
|
|
isLegacy: true,
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4-0314',
|
|
label: 'GPT-4 (0314)',
|
|
hidden: true, // OLD
|
|
description: 'Snapshot of gpt-4 from March 14th 2023 with function calling data. Data up to Sep 2021.',
|
|
contextWindow: 8192,
|
|
trainingDataCutoff: 'Sep 2021',
|
|
interfaces: [LLM_IF_OAI_Chat, LLM_IF_OAI_Fn],
|
|
chatPrice: { input: 30, output: 60 },
|
|
benchmark: { cbaElo: 1186 },
|
|
isLegacy: true,
|
|
},
|
|
{
|
|
idPrefix: 'gpt-4',
|
|
label: 'GPT-4',
|
|
symLink: 'gpt-4-0613',
|
|
isLegacy: true,
|
|
},
|
|
|
|
|
|
// 3.5
|
|
// Note: As of July 2024, gpt-4o-mini should be used in place of gpt-3.5-turbo, as it is cheaper, more capable, multimodal, and just as fast.
|
|
// As such, many 3.5 models are in the 'deny list' below, and not even returned to the UI.
|
|
{
|
|
idPrefix: 'gpt-3.5-turbo-0125',
|
|
label: '3.5-Turbo (2024-01-25)',
|
|
hidden: true, // OLD
|
|
description: 'The latest GPT-3.5 Turbo model with higher accuracy at responding in requested formats and a fix for a bug which caused a text encoding issue for non-English language function calls.',
|
|
contextWindow: 16385,
|
|
maxCompletionTokens: 4096,
|
|
trainingDataCutoff: 'Sep 2021',
|
|
interfaces: [LLM_IF_OAI_Chat, LLM_IF_OAI_Fn],
|
|
chatPrice: { input: 0.5, output: 1.5 },
|
|
benchmark: { cbaElo: 1106 },
|
|
},
|
|
{
|
|
idPrefix: 'gpt-3.5-turbo-1106',
|
|
label: '3.5-Turbo (1106)',
|
|
hidden: true, // OLD
|
|
description: 'GPT-3.5 Turbo model with improved instruction following, JSON mode, reproducible outputs, parallel function calling, and more.',
|
|
contextWindow: 16385,
|
|
maxCompletionTokens: 4096,
|
|
trainingDataCutoff: 'Sep 2021',
|
|
interfaces: [LLM_IF_OAI_Chat, LLM_IF_OAI_Fn],
|
|
chatPrice: { input: 1, output: 2 },
|
|
benchmark: { cbaElo: 1068 },
|
|
},
|
|
{
|
|
idPrefix: 'gpt-3.5-turbo',
|
|
label: '3.5-Turbo',
|
|
symLink: 'gpt-3.5-turbo-0125',
|
|
},
|
|
|
|
];
|
|
|
|
export const _fallbackOpenAIModel: KnownModel = {
|
|
idPrefix: '',
|
|
label: '?',
|
|
description: 'Unknown, please let us know the ID. Assuming a context window of 128k tokens, and a maximum output of 4k tokens.',
|
|
contextWindow: 128000,
|
|
maxCompletionTokens: 4096,
|
|
interfaces: [...IFS_CHAT_MIN, LLM_IF_OAI_Responses], // Use Responses API by default for new models
|
|
// hidden: true,
|
|
};
|
|
|
|
|
|
const openAIModelsDenyList: string[] = [
|
|
// [OpenAI, 2025-08-28] FIXME: NOT YET SUPPORTED - "REALTIME API"
|
|
// 'gpt-realtime', // leave this just for kicks, but it's hidden by default and won't work if unhidden
|
|
// '-realtime-',
|
|
'4o-realtime',
|
|
'4o-mini-realtime',
|
|
'gpt-realtime',
|
|
'gpt-realtime-mini',
|
|
|
|
// [OpenAI, 2025-03-11] FIXME: NOT YET SUPPORTED - "RESPONSES API"
|
|
'computer-use-preview', 'computer-use-preview-2025-03-11', // FIXME: support these
|
|
|
|
// [OpenAI Deprecations] Explicitly deny shut-down model IDs that we removed
|
|
// 'gpt-4.5-preview',
|
|
// 'o1-preview',
|
|
// 'gpt-4-32k',
|
|
|
|
// Legacy GPT models
|
|
'gpt-3.5-turbo-0301',
|
|
'gpt-3.5-turbo-0613',
|
|
'gpt-3.5-turbo-16k',
|
|
'gpt-3.5-turbo-16k-0613',
|
|
|
|
// Other unwanted GPT models
|
|
'gpt-4-turbo-preview',
|
|
|
|
// Non-chat GPT models
|
|
'-turbo-instruct', 'davinci-', 'babbage-',
|
|
|
|
// Embedding models: /v1/embeddings
|
|
'text-embedding-3-small', 'text-embedding-3-large', 'text-embedding-ada-002',
|
|
|
|
// TTS Models: /v1/audio/speech
|
|
'tts-1-hd', 'tts-1', 'gpt-4o-mini-tts', // FIXME: support these
|
|
|
|
// STT models: /v1/audio/transcriptions, /v1/audio/translations
|
|
'whisper-1', 'gpt-4o-transcribe', 'gpt-4o-mini-transcribe', // FIXME: support these
|
|
|
|
// Image models: /v1/images/generations
|
|
'gpt-image-1.5', 'chatgpt-image-latest', 'gpt-image-1', 'gpt-image-1-mini', 'dall-e-3', 'dall-e-2',
|
|
|
|
// Video models: /v1/videos
|
|
'sora-2-pro', 'sora-2',
|
|
|
|
// Moderation models
|
|
'omni-moderation-latest', 'omni-moderation-2024-09-26', 'text-moderation-latest',
|
|
];
|
|
|
|
export function openAIModelFilter(model: OpenAIWire_API_Models_List.Model) {
|
|
return !openAIModelsDenyList.some(deny => model.id.includes(deny));
|
|
}
|
|
|
|
export function openAIModelToModelDescription(modelId: string, modelCreated: number | undefined, modelUpdated?: number): ModelDescriptionSchema {
|
|
return fromManualMapping(_knownOpenAIChatModels, modelId, modelCreated, modelUpdated, _fallbackOpenAIModel);
|
|
}
|
|
|
|
export function openAIInjectVariants(acc: ModelDescriptionSchema[], model: ModelDescriptionSchema): ModelDescriptionSchema[] {
|
|
return createVariantInjector(hardcodedOpenAIVariants, 'after')(acc, model);
|
|
}
|
|
|
|
|
|
const _manualOrderingIdPrefixes = [
|
|
// GPT-5.2
|
|
'gpt-5.2-20',
|
|
'gpt-5.2-pro-20',
|
|
'gpt-5.2-pro',
|
|
'gpt-5.2-codex',
|
|
'gpt-5.2-chat-latest',
|
|
'gpt-5.2',
|
|
// GPT-5.1
|
|
'gpt-5.1-20',
|
|
'gpt-5.1-chat-latest',
|
|
'gpt-5.1-codex-max',
|
|
'gpt-5.1-codex',
|
|
'gpt-5.1-codex-mini',
|
|
'gpt-5.1',
|
|
// GPT-5
|
|
'gpt-5-20',
|
|
'gpt-5-pro-20',
|
|
'gpt-5-pro',
|
|
'gpt-5-mini-20',
|
|
'gpt-5-nano-20',
|
|
'gpt-5-chat-latest',
|
|
'gpt-5-codex',
|
|
'gpt-5-',
|
|
// OSB models?
|
|
'osb-',
|
|
// Reasoning models
|
|
'o5-20',
|
|
'o5-mini-20',
|
|
'o5-',
|
|
'o4-pro-20',
|
|
'o4-pro',
|
|
'o4-mini-deep-research-20',
|
|
'o4-mini-deep-research',
|
|
'o4-20',
|
|
'o4-mini-20',
|
|
'o4-',
|
|
'o3-pro-20',
|
|
'o3-pro',
|
|
'o3-deep-research-20',
|
|
'o3-deep-research',
|
|
'o3-20',
|
|
'o3-mini-20',
|
|
'o3-',
|
|
'o1-pro-20',
|
|
'o1-pro',
|
|
'o1-20',
|
|
'o1-preview-',
|
|
'o1-',
|
|
// GPT-4.5
|
|
'gpt-4.5-20',
|
|
'gpt-4.5-preview',
|
|
'gpt-4.5-',
|
|
// GPT-4.1
|
|
'gpt-4.1-20',
|
|
'gpt-4.1-mini-20',
|
|
'gpt-4.1-mini',
|
|
'gpt-4.1-nano-20',
|
|
'gpt-4.1-nano',
|
|
'gpt-4.1',
|
|
// 4o-derived?
|
|
'gpt-audio-2',
|
|
'gpt-4o-audio-preview',
|
|
'gpt-audio-mini-',
|
|
'gpt-audio-mini',
|
|
'gpt-4o-mini-audio-preview',
|
|
'gpt-audio',
|
|
// Preferred models
|
|
'gpt-4o-20',
|
|
'gpt-4o-search-20',
|
|
'gpt-4o-search-preview',
|
|
'gpt-4o-mini-20',
|
|
'gpt-4o-mini-search-20',
|
|
'gpt-4o-mini-search-preview',
|
|
'gpt-4o-mini',
|
|
'gpt-4o-',
|
|
// ChatGPT models
|
|
'chatgpt-',
|
|
// Codex
|
|
'codex-',
|
|
// Computer use models
|
|
'computer-use-20',
|
|
'computer-use-preview',
|
|
'computer-use',
|
|
// ...rest
|
|
// 'gpt-4-turbo-',
|
|
// 'gpt-4-',
|
|
// ...
|
|
];
|
|
|
|
|
|
export function openAISortModels(a: ModelDescriptionSchema, b: ModelDescriptionSchema) {
|
|
// bottom: links
|
|
const aLink = a.label.startsWith('🔗');
|
|
const bLink = b.label.startsWith('🔗');
|
|
if (aLink !== bLink) return aLink ? 1 : -1;
|
|
|
|
// bottom: non-chatGenerate
|
|
const aChat = a.interfaces.includes(LLM_IF_OAI_Chat);
|
|
const bChat = b.interfaces.includes(LLM_IF_OAI_Chat);
|
|
if (aChat !== bChat) return aChat ? -1 : 1;
|
|
|
|
// sort by manual ordering (if not present is implicitly at the bottom)
|
|
const aOrder = _manualOrderingIdPrefixes.findIndex(prefix => a.id.startsWith(prefix));
|
|
const bOrder = _manualOrderingIdPrefixes.findIndex(prefix => b.id.startsWith(prefix));
|
|
if (aOrder !== bOrder) {
|
|
if (aOrder === -1) return 1;
|
|
if (bOrder === -1) return -1;
|
|
return aOrder - bOrder;
|
|
}
|
|
|
|
// fix the OpenAI model names to be chronologically sorted
|
|
function remapReleaseDate(id: string): string {
|
|
return id
|
|
.replace('0314', '2023-03-14')
|
|
.replace('0613', '2023-06-13')
|
|
.replace('1106', '2023-11-06')
|
|
.replace('0125', '2024-01-25')
|
|
.replace('0409', '2024-04-09');
|
|
}
|
|
|
|
// due to using by-label, sorting doesn't require special cases anymore
|
|
return remapReleaseDate(b.label).localeCompare(remapReleaseDate(a.label));
|
|
}
|
|
|
|
|
|
/**
|
|
* Checks for model definition issues:
|
|
* 1. Stale definitions: in our known models but not in API (should remove)
|
|
* 2. Unknown models: in API but not in our known models (should add)
|
|
* 3. Parsing gaps: in API but lost during parsing pipeline (bug detection)
|
|
*/
|
|
export function openaiValidateModelDefs_DEV(apiModels: unknown, parsedModels?: object[]): void {
|
|
|
|
if (DEV_DEBUG_OPENAI_MODELS) {
|
|
|
|
// Check if wireModels has .data array
|
|
if (!apiModels || !Array.isArray(apiModels)) {
|
|
console.warn('[DEV] OpenAI: wireModels is not an array', apiModels);
|
|
return;
|
|
}
|
|
|
|
const apiIds = apiModels.map((model: any) => model.id);
|
|
const knownIds = _knownOpenAIChatModels
|
|
.filter(model => model.idPrefix && model.idPrefix !== '') // exclude fallback model
|
|
.map(model => model.idPrefix);
|
|
|
|
// 1 & 2: Check stale and unknown definitions
|
|
llmDevCheckModels_DEV('OpenAI', apiIds, knownIds, { apiFilter: id => openAIModelFilter({ id } as any) });
|
|
|
|
// 3: Check for parsing gaps (models lost during filter/map pipeline)
|
|
if (parsedModels) {
|
|
const parsedIds = new Set(parsedModels.map((m: any) => m.id));
|
|
const parsingGaps = apiIds.filter(id =>
|
|
!parsedIds.has(id) && // not in parsed output
|
|
openAIModelFilter({ id } as any), // not filtered by deny list
|
|
);
|
|
if (parsingGaps.length)
|
|
console.log(`[DEV] OpenAI: parsing gaps (in API, passed filter, but not in output): [ ${parsingGaps.join(', ')} ]`);
|
|
}
|
|
|
|
}
|
|
|
|
}
|