Devstral-2-123B-Instruct-2512 vs GPT-4o (2024-11-20)

API pricing, benchmark indices and specs — updated 19 October 2026.

mistralai · A · Open weights

Params

125.0 B

Context

262,144

30d

5 K

openai · B · API-only

Input / 1M

$2.50

Output / 1M

$10.00

AA index

8.4

proprietary · 128K context

Specification comparison

Differences are highlighted; identical values are muted.

Specification	Devstral-2-123B-Instruct-2512	GPT-4o (2024-11-20)
Deployment	Open weights · also via API	API-only (closed weights)
Parameters	125.0 B	n/a — API-only
Architecture	Ministral3ForCausalLM	n/a — API-only
Context length	262,144	128,000
Licence	other	Proprietary
Commercial use	Unknown	Per API terms
Languages	—	n/a — API-only
Downloads 30d	5,315	n/a — API-only
Downloads all time	353 K	n/a — API-only
Quantizations on the Hub	SAFETENSORS	n/a — API-only
Gated	No	n/a — API-only
First seen on the Hub	2025-11-28	n/a — API-only
API input price / 1M tokens	$0.40	$2.50
API output price / 1M tokens	$2.00	$10.00
Est. cost per 1M input + 1M output tokens	$2.40	$12.50
Cached input / 1M tokens	$0.04	$1.25
AA Intelligence Index	8.6	8.4
AA Coding Index	31.3	—
AA Agentic Index	2.3	—
Max output tokens	209,715	16,384
API providers	1	1
Tool calling	Yes	Yes
Reasoning mode	No	No

Estimated cost of 1M input + 1M output tokens via OpenRouter

Devstral-2-123B-Instruct-2512 $2.40

GPT-4o (2024-11-20) $12.50

Devstral-2-123B-Instruct-2512 is 5.2× cheaper at list price.

On RTX 4090 · 24 GB · 8K context unless noted

Devstral-2-123B-Instruct-2512 · fp16 @ 256K ctx 882.8 GB / 24 GB

❌ Won’t fit