Mistral Large 3 vs Kimi K3

Mistral AI's Mistral Large 3 and Moonshot AI's Kimi K3 compared on specs, pricing, and capabilities — as configured in AI Crucible ensembles.

Mistral Large 3 input tokens are 6.0x cheaper than Kimi K3 ($0.60 vs $3.60 per 1M tokens).
Mistral Large 3 output tokens are 10.0x cheaper than Kimi K3 ($1.80 vs $18.00 per 1M tokens).
Kimi K3 offers the larger context window: 1.048576M tokens vs 256K tokens for Mistral Large 3.
Kimi K3 is a reasoning model that spends extra tokens thinking before answering; Mistral Large 3 responds directly.

Specification	Mistral Large 3	Kimi K3
Provider	Mistral AI	Moonshot AI
API model ID	mistral-large-latest	kimi-k3
Description	Mistral AI's flagship model with 41B active parameters, excelling at complex reasoning and multimodal tasks.	Moonshot AI's 2.8T-parameter flagship for long-horizon coding and knowledge work, with always-on reasoning, vision, and 1M context.
Context window	256K tokens	1.048576M tokens
Max output tokens	8,192	8,192
Input cost ($/1M tokens)	$0.60	$3.60
Output cost ($/1M tokens)	$1.80	$18.00
Cache read cost ($/1M tokens)	-	$0.36
Latency	medium	high
Reasoning model	No	Yes
Vision (image input)	Yes	Yes
Tool use	Yes	Yes