How does API pricing compare between FLUX.2 Dev and Mistral Large 4?

FLUX.2 Dev costs $0.03/M input / $0.03/M output per 1M tokens. Mistral Large 4 costs $2/M input / $6/M output per 1M tokens. FLUX.2 Dev is the cheaper option.

FLUX.2 DevvsMistral Large 4

Q: Which model should I choose for my use case?

Choose FLUX.2 Dev for vision and multimodal apps, and Mistral Large 4 for knowledge and factual QA.

FLUX.2 Dev leads the overall Podium Score by 0.6 points (90.4 vs 89.8).Category wins: FLUX.2 Dev 1 — 6 Mistral Large 4.FLUX.2 Dev is the cheaper pick ($0.03/M vs $6/M per 1M output tokens). Mistral Large 4 is faster (92 t/s vs 28 t/s).

Open in Compare Tool →View FLUX.2 Dev View Mistral Large 4

FLUX.2 DevDrag slider or hover to compareMistral Large 4

Mistral Large 4

89.8

Coding: 84.5 · Speed: 92 t/s

Output Price: $6/M

FLUX.2 Dev

90.4

Coding: 0.0 · Speed: 28 t/s

Output Price: $0.03/M

Quick VerdictAggregated 2026 Benchmark Analysis

FLUX.2 Dev is rated higher overall with a Podium Score of 90.4 (vs 89.8 for Mistral Large 4). In head-to-head domain evaluations, Mistral Large 4 outperforms in 6 of 7 benchmark categories.

Coding & Engineering

Mistral Large 4 (84.5)

Inference Speed

Mistral Large 4 (92 t/s)

Cost Efficiency

FLUX.2 Dev ($0.03/M)

API Price / 1M (in · out)

FLUX.2 Dev $0.03/M · $0.03/M — Mistral Large 4 $2/M · $6/M

Choose FLUX.2 Dev for vision and multimodal apps, and Mistral Large 4 for knowledge and factual QA.

Metric	FLUX.2 Dev	Mistral Large 4
Podium Score	90.4 ▲	89.8
Arena Elo	1440	1450 ▲
Intelligence Index	90.0	90.0
Coding	0.0	84.5 ▲
Math	0.0	87.0 ▲
Reasoning	42.0	89.5 ▲
Agentic	28.0	81.0 ▲
Knowledge	60.0	91.0 ▲
Multimodal	98.5 ▲	90.0
Long-context	55.0	88.0 ▲
Output speed	28 t/s	92 t/s ▲
Output price	$0.03/M ▲	$6/M
Context window	128K	256K ▲

Scores aggregated from five independent leaderboards. See Methodology. More head-to-heads: FLUX.2 Dev vs Claude Mythos · FLUX.2 Dev vs Claude 5.5 Sonnet · FLUX.2 Dev vs Kimi K3.1 Pro · FLUX.2 Dev vs Eleven Multilingual v3