Roboflow

Gemini 4 Neon vs Mistral Large 4

Gemini 4 Neon and Mistral Large 4 score 91.0% vs 56.8% averaged across five Vision Evals tasks (#1 and #51 of 73 models), so Gemini 4 Neon leads overall.

  • Gemini 4 Neon wins all 5 tasks, with its widest lead on reasoning (92.9% vs 38.9%).
  • Mistral Large 4 is about 1/7 of the cost: $0.0033 vs $0.022 overall average cost per sample.
  • Gemini 4 Neon is faster: 20.7s vs 31.2s per sample.

Compare Gemini 4 Neon vs Mistral Large 4 live

Run the same image across every model that supports a task and compare their outputs side-by-side.

Detect and compare bounding boxes across models on the same image.

Open Object Detection in the full playground
GoogleGemini 4 Neon
Run to compare this model.
MistralMistral Large 4
Run to compare this model.

Gemini 4 Neon vs Mistral Large 4 on Vision Evals

Score by taskGemini 4 NeonMistral Large 4
Reasoning
92.9
38.9
OCR
88.1
46.1
Counting
92.3
54.5
Object Detection
84.5
59.3
Identification
96.9
85.4

Roboflow Vision Evals. Full breakdown below.

Gemini 4 Neon vs Mistral Large 4 Comparison Table

Evals updated November 19, 2026Pricing updated November 19, 2026

PropertyGemini 4 NeonMistral Large 4
OrganizationGoogleMistral
Categoryclosedopen
Modalitymultimodalmultimodal
Release DateNov 2026Oct 2026
Context Window—1.0M
ParametersUnknown1.05T total, 49B active
LicenseProprietaryCustom
Commercial usecommercial use per vendor termsSee license terms
Pricing per 1M tokens
Input $/1MNo published price$0.700
Output $/1MNo published price$2.10
Vision Tasks
Chart Question AnsweringNot listedSupported
Document Question AnsweringNot listedSupported
Image TaggingNot listedSupported
Multi-Label ClassificationNot listedSupported
Phrase GroundingNot listedSupported
Promptable Concept SegmentationDemoNot listed
Both support 6 more tasks
Model Features
Both support 3 more features
Vision Evalsground-truth scores across 5 vision tasks, pooled at low effort
Overall
91.0%
56.8%
Avg cost / sample$0.022$0.0033
Avg speed / sample20.70s31.23s
By task
Object Detection (low)
84.5%
±0.4, Mean of 3 runs, range 84.0 to 84.8
$0.035
59.3%
±0.7, Mean of 3 runs, range 58.5 to 60.0
$0.0028
Object Detection (high)
85.8%
±0.5, Mean of 3 runs, range 85.3 to 86.2
$0.062
50.2%
±2.5, Mean of 3 runs, range 48.0 to 53.0
$0.023
Counting (low)
92.3%
±0.7, Mean of 3 runs, range 91.9 to 93.2
$0.011
54.5%
±0.7, Mean of 3 runs, range 54.0 to 55.4
$0.0011
Counting (high)
89.2%
±1.4, Mean of 3 runs, range 87.8 to 90.5
$0.055
63.1%
±2.0, Mean of 3 runs, range 60.8 to 64.9
$0.0095
Identification (low)
96.9%
±0.0, Mean of 3 runs, range 96.9 to 96.9
$0.0064
85.4%
±3.1, Mean of 3 runs, range 81.3 to 87.5
$0.0009
Identification (high)
96.9%
±0.0, Mean of 3 runs, range 96.9 to 96.9
$0.010
85.4%
±1.6, Mean of 3 runs, range 84.4 to 87.5
$0.0044
OCR (low)
88.1%
$0.021
46.1%
$0.0044
by category
Single value
80.9%
Transcription
98.5%
Structured JSON
93.8%
Text localization
86.3%
Single value
33.5%
Transcription
68.9%
Structured JSON
71.7%
Text localization
0.0%
OCR (high)
89.2%
$0.082
54.5%
$0.021
by category
Single value
83.5%
Transcription
98.8%
Structured JSON
93.9%
Text localization
86.0%
Single value
38.3%
Transcription
75.3%
Structured JSON
76.4%
Text localization
28.8%
Reasoning (low)
92.9%
±0.3, Mean of 3 runs, range 92.7 to 93.4
$0.011
38.9%
±0.3, Mean of 3 runs, range 38.4 to 39.1
$0.0013
Reasoning (high)
94.0%
±0.7, Mean of 3 runs, range 93.4 to 94.7
$0.038
57.6%
±2.0, Mean of 3 runs, range 55.6 to 59.6
$0.013

Gemini 4 Neon vs Mistral Large 4: Overview

Gemini 4 Neon

Gemini 4 Neon is a multimodal model from Google. In Roboflow Playground, it supports image captioning, OCR, classification, object detection, visual question answering and instance segmentation through the native Gemini workflow block. Compare its low- and high-thinking results on the current VLM Exam benchmark, or run it on your own images.

Mistral Large 4

Mistral Large 4, nicknamed Le Chonk, is a natively multimodal mixture-of-experts model from Mistral that accepts interleaved text and image input and produces text output. It uses a granular MoE design with roughly 1.05 trillion total parameters and 49 billion active per token, reported as 52 billion when embeddings and output layers are counted, paired with a 1.6 billion parameter vision encoder and a context window of one million tokens. The model is trained from scratch on about 3,800 NVIDIA Grace Blackwell GPUs in Mistral's European data centers and supports more than 160 languages. It behaves as a hybrid instruct and reasoning system, with a reasoning effort setting that selects between direct answers and longer deliberation, alongside function calling and structured output for agentic workflows.

Image understanding is a focus of this generation, covering documents, charts, technical drawings and natural scenes, and the model emits bounding box coordinates for visual grounding queries. Reported grounding results include 42 percent on Dense200 and 73 percent on the DIOR-RSVG remote sensing benchmark. Mistral describes agentic vision workflows in which the model zooms into gigapixel satellite imagery or engineering drawings to verify details, and reports coding results such as 62 percent on DeepSWE. Figures published at preview time are preliminary because the reinforcement learning phase is still in progress.

Frequently Asked Questions

On Roboflow's Vision Evals, Gemini 4 Neon performed better. It scores higher on all five vision tasks and averages 91.0% (#1 of 73) against 56.8% (#51 of 73) for Mistral Large 4. These scores measure vision capabilities only; either model may perform differently on work outside this benchmark's tasks.

Yes. On the Vision Evals Reasoning benchmark at low effort, Gemini 4 Neon leads with 92.9% against 38.9%. This is the widest gap between the two models across the benchmark's tasks.

Mistral Large 4 is cheaper to run. Across Roboflow's Vision Evals task mix it averaged $0.0033 per sample against $0.022. Actual costs depend on your image sizes, prompts, and output length.

Gemini 4 Neon is faster. Across Roboflow's Vision Evals it averaged 20.7s per inference against 31.2s. Latency varies with image size, prompt length, and provider load, so treat these as relative rather than guaranteed figures.

Yes. The comparison demo on this page runs both models on the same image side by side for image captioning and OCR in the free Roboflow Playground. You can try it instantly, and a free account gives you more runs with free Roboflow credits.