The best LLMs for your use case:

1Kimi K2.5Moonshot

1T-parameter MoE reasoning model with state-of-the-art performance on math, code, and multimodal tasks.

Speed:

Intelligence:

Price: (1M Tokens)

$0.50 / 2.80

Inputs:

ImageText

JSON Mode:

Function Calling:

Benchmarks:

#4

SimpleQA

General Knowledge

36.9
#1

MMLU-Pro

General Knowledge

87.1
#1

GPQA-Diamond

General Knowledge

87.6
#1

LMArena

Chat

1447
#1

LiveCodeBench

Code

85
#1

WebDevArena

Code

1446
#1

LongBenchv2

Summarization

61
#1

MMMU

Multimodal - Vision

84.3
#4

SimpleQA

General Knowledge

36.9
#1

MMLU-Pro

General Knowledge

87.1
#1

GPQA-Diamond

General Knowledge

87.6
#1

LMArena

Chat

1447
#1

LiveCodeBench

Code

85
#1

WebDevArena

Code

1446
#1

LongBenchv2

Summarization

61
#1

MMMU

Multimodal - Vision

84.3
Try it out
2DeepSeek-V3.1DeepSeek

Advanced Mixture-of-Experts model with improved efficiency and performance over V3, delivering enhanced reasoning capabilities at competitive cost.

Speed:

Intelligence:

Price: (1M Tokens)

$0.71

Inputs:

ImageText

JSON Mode:

Function Calling:

Benchmarks:

#5

GPQA-Diamond

General Knowledge

80.1
#3

MMLU-Pro

General Knowledge

84.8
#1

SimpleQA

General Knowledge

93.4
#5

LMArena

Chat

1419
#5

LiveCodeBench

Code

74.8
#5

WebDevArena

Code

1364
#5

GPQA-Diamond

General Knowledge

80.1
#3

MMLU-Pro

General Knowledge

84.8
#1

SimpleQA

General Knowledge

93.4
#5

LMArena

Chat

1419
#5

LiveCodeBench

Code

74.8
#5

WebDevArena

Code

1364
Try it out

Use case:

General Knowledge