Home/Use Cases/Reasoning

Best LLM for Reasoning

8 models ranked for reasoning tasks. Sorted by benchmark quality score, with price as a secondary factor.

Best Quality

OpenAI

ELO 1420

Cheapest Option

DeepSeek

$0.28/1M/1M input

All Models for Reasoning

#	Model	Provider	Input / 1M	Output / 1M	ELO	Flags
🥇	GPT-5.4	OpenAI	$2.50/1M	$15.00/1M	1420
🥈	Claude Opus 4.7	Anthropic	$5.00/1M	$25.00/1M	1415
🥉	Gemini 3.1 Pro	Google	$2.00/1M	$12.00/1M	1410
4	o3	OpenAI	$2.00/1M	$8.00/1M	1395
5	Gemini 2.5 Pro	Google	$1.25/1M	$10.00/1M	1385
6	DeepSeek V3.2 (Reasoner)	DeepSeek	$0.28/1M	$0.42/1M	1385	OSS
7	Grok 4.20	xAI	$2.00/1M	$6.00/1M	1380
8	o4-mini	OpenAI	$1.10/1M	$4.40/1M	1350

$2.50/1M/1MELO 1420

GPT-5. OpenAI flagship — top-tier reasoning, coding, and multimodal.

$5.00/1M/1MELO 1415

Claude Opus 4. Step-change improvement in agentic coding over Opus 4.6.

$2.00/1M/1MELO 1410

Gemini 3. Google's latest flagship with major reasoning improvements.

GPT-5.4 by OpenAI is rated as the best model for reasoning with an ELO score of 1420. OpenAI flagship — top-tier reasoning, coding, and multimodal.

DeepSeek V3.2 (Reasoner) is the most affordable option for reasoning at $0.28/1M per 1M input tokens.

No completely free models are listed for reasoning, but DeepSeek V3.2 (Reasoner) start at very low prices.