Inkling (Xhigh)
Thinking MachinesOpen WeightApache 2.0 · Uso Comercial
Descripción
Inkling is a general-purpose multimodal model that accepts text, image and audio inputs and generates text outputs.
Fecha de lanzamiento
2026-07-15
Parámetros
975.0B
Longitud del contexto
262K
Modalidades
audio, image, text
Radar de capacidades
26
general
51
coding
87
reasoning
62
science
50
agents
80
multimodal
Rankings
| Dominio | #Posición | Puntuación | Fuente |
|---|---|---|---|
| Capacidad agéntica | 40 | 52.0 | LS |
| Ranking de codificación | 108 | 72.0 | AA |
| Ranking general | 261 | 40.0 | AA |
| Razonamiento matemático | 37 | 88.0 | LB |
| Ranking multimodal | 90 | 49.0 | LS |
| Razonamiento | 43 | 78.0 | LB |
| Ciencia | 88 | 69.0 | AA |
Puntuaciones de benchmarks (LLM Stats)
(LLM Stats (zeroeval))Agents
Tau3 Banking
23.7%Aut.
Audio
MMAU
77.2%Aut.
Chat
VoiceBench Avg
91.4%Aut.
Factuality
SimpleQA Verified
43.9%Aut.
General
GDPval-AA
1238.00 / 3000Aut.
Global-MMLU-Lite
88.7%Aut.
Instruction Following
IFBench
79.8%Aut.
Knowledge
AA-Omniscience Index
2.10 / 200Aut.
Math
AIME 2026
97.1%Aut.
Reasoning
CharXiv-R
78.1%Aut.
SWE-Bench Verified
77.6%Aut.
MCP Atlas
76.0%Aut.
Terminal-Bench 2.1
63.8%Aut.
Humanity's Last Exam (with tools, text-only)
46.0%Aut.
Humanity's Last Exam (no tools, text-only)
29.7%Aut.
Vision
MMMU-Pro
73.5%Aut.
Índices de evaluación AA
(Artificial Analysis)Gpqa(NYU + Cohere + Anthropic (2023))87.2
Lcr(Artificial Analysis)77.3
Terminalbench V2 155.1
Coding Index(Artificial Analysis)52.1
Scicode(UIUC + Argonne National Lab (2024))47.0
Hle(Center for AI Safety + Scale AI (2025))31.9
Tau Banking29.1
Intelligence Index(Artificial Analysis)25.0
Terminalbench V4 01.0
Puntuaciones por categoría LLM Stats
(LLM Stats (zeroeval))Legal100
Finance100
Agents100
Reasoning83
General78
Knowledge2
Science2
Chat90
Language90
Safety90
Speech To Text90
Communication90
Instruction Following80
Multimodal80
Frontend Development80
Audio80
Vision80
Code70
Math60
Tool Calling50
Factuality40
Precios
Precio de entrada$1 / 1M tokens
Precio de salida$4.05 / 1M tokens
Precio mixto (3:1)$1.762 / 1M tokens
Precio de lectura caché$0.374 / 1M tokens
Velocidad
Tokens/seg170.9
Retraso del primer token2.02s
Tiempo hasta la respuesta13.72s
Ranking de Precios por Proveedor
Ranking de Precios por Proveedor
18 proveedores
Más barato: DeepInfraMás caro: Abacus
ProveedorEntradaSalida
1DeepInfraMás barato
$0
$0
2Venice AI
$0.5625
$1.5
3Deep Infra
$0.95
$4.05
4Thinking MachinesPRINCIPAL
$1
$4.05
5NanoGPT
$1
$4.05
6Hugging Face
$1
$4.05
7Kilo Gateway
$1
$4.05
8Vercel AI Gateway
$1
$4.05
9Together AI
$1
$4.05
10Merge Gateway
$1
$4.05
11Neon
$1
$4.05
12CommandCode
$1
$4.05
13AI/ML API
$1.058
$4.2849
14Charm Hyper
$1.0888
$4.40964
15Requesty
$1.87
$4.68
16LLMTR
$1.87
$4.68
17Impossibl
$1.87
$4.68
18Abacus
$3.74
$9.36
Comparar precios entre diferentes proveedores de API para este modelo.