LLM models
Compare LLM Models
Compare LLM Models
Compare current language models across sourced capability scores, runtime, pricing, and catalog metadata.
575 modelsUpdated Jul 16, 2026, 12:22 PM
Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)
GPT-5.6 Sol (max)
GPT-5.6 Sol (xhigh)
GPT-5.6 Sol (max)
OpenAI
GPT-5.6 Sol (xhigh)
OpenAI
Capability
Overall
59.9
#1 of 566
58.9
#2 of 566
57.7
#3 of 566
Coding
76.5
#5 of 222
77.4
#2 of 222
78.3
#1 of 222
Math
n/a
Not published by available benchmark sources
n/a
Not published by available benchmark sources
n/a
Not published by available benchmark sources
LiveBench Math
96.0%
#2 of 26 · LiveBench 2026-06-25
96.2%
#1 of 26 · LiveBench 2026-06-25
95.5%
#4 of 26 · LiveBench 2026-06-25
MMLU-Pro
n/a
Not published by available benchmark sources
n/a
Not published by available benchmark sources
n/a
Not published by available benchmark sources
Reasoning
99.5
#3 of 543
99.6
#2 of 543
98.9
#6 of 543
Runtime & Price
Output Speed
69.4 tok/s
#201 of 314
67.4 tok/s
#205 of 314
57 tok/s
#239 of 314
First Token
83.88s
#307 of 314
92.43s
#309 of 314
26.47s
#295 of 314
Blended Price
$20.00/M
#360 of 371
$11.25/M
#346 of 371
$11.25/M
#347 of 371
Input Price
$10.00/M
#359 of 371
$5.00/M
#337 of 371
$5.00/M
#338 of 371
Output Price
$50.00/M
#360 of 371
$30.00/M
#346 of 371
$30.00/M
#347 of 371
Value Index
3.0
#334 of 369
5.2
#303 of 369
5.1
#304 of 369
Metadata
Creator
Anthropic
OpenAI
OpenAI
Type
n/a
n/a
n/a
Release
Jun 9, 2026
Jul 9, 2026
Jul 9, 2026
Context
1,000,000
1,050,000
1,050,000
Max Output
128,000
128,000
128,000
Modalities
In: text, image, file / Out: text
In: file, image, text / Out: text
In: file, image, text / Out: text
Model Traits
n/a
n/a
n/a
Direct benchmark values come from Artificial Analysis, the public TIGER-Lab MMLU-Pro leaderboard, and LiveBench. Imported values retain their source, and missing scores are never inferred. OpenRouter and Vercel AI Gateway enrich matched model metadata; Value Index is derived inside Easy Benchmarks.