vs
2:1
overall capability wins (0 ties across 3 scored). Ministral 3 3B leads overall.
- Pricing3:0
- Speed2:0
- Specs2:1
Verdict
Ministral 3 3B offers ~2.0x larger context (256K vs 128K); Ministral 3 3B allows ~4.0x more max output (33K vs 8K); Ministral 3 3B is ~1.5x cheaper on a blended token basis than Mistral Nemo 12B.
- Ministral 3 3B offers ~2.0x larger context (256K vs 128K)
- Ministral 3 3B allows ~4.0x more max output (33K vs 8K)
- Ministral 3 3B is ~1.5x cheaper on a blended token basis than Mistral Nemo 12B
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- Ministral 3 3B
- $0.13
- Mistral Nemo 12B
- $0.19
Ministral 3 3B is about 1.5x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
Mistral
Mistral
Family
Mistral
Mistral
License
Apache 2.0
Apache 2.0
Open weights
Yes
Yes
Release
Dec 1, 2025
Jul 18, 2024
Knowledge cutoff
-
-
API / provider
Mistral
Mistral
Modalities
text, image → text
text → text
Specs
Context window
256K
128K
Max output
33K
8K
Parameters
3B
12B
Pricing
Input $/1M
$0.10
$0.15
Output $/1M
$0.10
$0.15
Blended $/1M (3∶1)
$0.10
$0.15
Speed
tok/s
180
140
TTFT (s)
0.15
0.2
Reasoning
MMLU-Pro
52.4
—
GPQA Diamond
35.8
—
Humanity's Last Exam
5.4
—
AIME 2025
22.0
—
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
9.0
—
AA-LCR v1.1
17.0
—
CritPt
—
—
MMMU-Pro
38.1
—
IFBench
26.8
—
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
24.7
—
Terminal-Bench 2.1
0.0
—
Aider Polyglot
—
—
Terminal-Bench 3
—
—
BigCodeBench
—
—
SciCode
15.3
—
CursorBench
—
—
SWE-Rebench
—
—
NL2Repo-Bench
—
—
DeepSWE
—
—
WebDev Arena
—
—
Terminal-Bench 4.0
—
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
—
—
| Metric | Ministral 3 3B | Mistral Nemo 12B | Delta |
|---|---|---|---|
| Identity | |||
| Organization | Mistral | Mistral | — |
| Family | Mistral | Mistral | — |
| License | Apache 2.0 | Apache 2.0 | — |
| Open weights | Yes | Yes | — |
| Release | Dec 1, 2025 | Jul 18, 2024 | — |
| Knowledge cutoff | - | - | — |
| API / provider | Mistral | Mistral | — |
| Modalities | text, image → text | text → text | — |
| Specs | |||
| Context window | 256K | 128K | A +128K |
| Max output | 33K | 8K | A +25K |
| Parameters | 3B | 12B | B +9.00 |
| Pricing | |||
| Input $/1M | $0.10 | $0.15 | A +$0.05 |
| Output $/1M | $0.10 | $0.15 | A +$0.05 |
| Blended $/1M (3∶1) | $0.10 | $0.15 | A +$0.05 |
| Speed | |||
| tok/s | 180 | 140 | A +40 |
| TTFT (s) | 0.15 | 0.2 | A +0.05 |
| Reasoning | |||
| MMLU-Pro | 52.4 | — | — |
| GPQA Diamond | 35.8 | — | — |
| Humanity's Last Exam | 5.4 | — | — |
| AIME 2025 | 22.0 | — | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 9.0 | — | — |
| AA-LCR v1.1 | 17.0 | — | — |
| CritPt | — | — | — |
| MMMU-Pro | 38.1 | — | — |
| IFBench | 26.8 | — | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | 24.7 | — | — |
| Terminal-Bench 2.1 | 0.0 | — | — |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | — | — |
| SciCode | 15.3 | — | — |
| CursorBench | — | — | — |
| SWE-Rebench | — | — | — |
| NL2Repo-Bench | — | — | — |
| DeepSWE | — | — | — |
| WebDev Arena | — | — | — |
| Terminal-Bench 4.0 | — | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | — | — | — |
Ministral 3 3B
- Context
- 256K / 33K out
- Parameters
- 3B
- Price
- $0.10 / $0.10
- Speed
- 180 tok/s · 0.15s TTFT
- Modalities
- text, image → text
- License
- Apache 2.0
Tiny Ministral 3 multimodal model - cheapest Mistral hosted tier for edge and classification.
Mistral Nemo 12B
- Context
- 128K / 8K out
- Parameters
- 12B
- Price
- $0.15 / $0.15
- Speed
- 140 tok/s · 0.2s TTFT
- Modalities
- text → text
- License
- Apache 2.0
Compact open Mistral x NVIDIA collab model - popular edge and multilingual baseline.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.
These two models have no overlapping published benchmarks in our dataset. Compare specs and pricing instead.