Mistral Large 3 vs DeepSeek-V4-Pro

Mistral AI's Mistral Large 3 and DeepSeek's DeepSeek-V4-Pro compared on specs, pricing, and capabilities — as configured in AI Crucible ensembles.

DeepSeek-V4-Pro input tokens are 1.1x cheaper than Mistral Large 3 ($0.52 vs $0.60 per 1M tokens).
DeepSeek-V4-Pro output tokens are 1.7x cheaper than Mistral Large 3 ($1.04 vs $1.80 per 1M tokens).
DeepSeek-V4-Pro offers the larger context window: 1M tokens vs 256K tokens for Mistral Large 3.
DeepSeek-V4-Pro is a reasoning model that spends extra tokens thinking before answering; Mistral Large 3 responds directly.
Mistral Large 3 accepts image inputs; DeepSeek-V4-Pro is text-only in AI Crucible.

Specification	Mistral Large 3	DeepSeek-V4-Pro
Provider	Mistral AI	DeepSeek
API model ID	mistral-large-latest	deepseek-v4-pro
Description	Mistral AI's flagship model with 41B active parameters, excelling at complex reasoning and multimodal tasks.	DeepSeek's most capable V4 model with dual thinking/non-thinking modes and 1M context.
Context window	256K tokens	1M tokens
Max output tokens	8,192	8,192
Input cost ($/1M tokens)	$0.60	$0.52
Output cost ($/1M tokens)	$1.80	$1.04
Cache read cost ($/1M tokens)	-	$0.00
Latency	medium	medium
Reasoning model	No	Yes
Vision (image input)	Yes	No
Tool use	Yes	Yes