Roboflow

MiMo V2.6 Pro vs Mistral Large 4

Mistral Large 4 and MiMo V2.6 Pro score 56.8% vs 50.3% averaged across five Vision Evals tasks (#46 and #59 of 68 models), so Mistral Large 4 leads overall.

  • Mistral Large 4 wins 4 of the 5 tasks, with its widest lead on object detection (59.3% vs 42.0%).
  • MiMo V2.6 Pro is about a third of the cost: $0.0011 vs $0.0033 overall average cost per sample.
  • MiMo V2.6 Pro is faster: 9.5s vs 31.2s per sample.

Compare MiMo V2.6 Pro 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
MiMo V2.6 Pro
Run to compare this model.
MistralMistral Large 4
Run to compare this model.

MiMo V2.6 Pro vs Mistral Large 4 on Vision Evals

Score by taskMiMo V2.6 ProMistral Large 4
Object Detection
42.0
59.3
Identification
76.0
85.4
Counting
50.0
54.5
Reasoning
35.1
38.9
OCR
48.6
46.1

Roboflow Vision Evals. Full breakdown below.

MiMo V2.6 Pro vs Mistral Large 4 Comparison Table

Evals updated November 10, 2026Pricing updated November 12, 2026

PropertyMiMo V2.6 ProMistral Large 4
OrganizationXiaomiMistral
Categoryopenopen
Modalitymultimodalmultimodal
Release DateSep 2026Oct 2026
Context Window1.0M1.0M
Parameters1.02T total, 42B active1.05T total, 49B active
LicenseMITCustom
Commercial usefree for commercial useSee license terms
Pricing per 1M tokens
Input $/1M$0.435$0.700
Output $/1M$0.870$2.10
Vision Tasks
Phrase GroundingNot listedSupported
Both support 10 more tasks
Model Features
Both support 3 more features
Vision Evalsground-truth scores across 5 vision tasks, pooled at low effort
Overall
50.3%
56.8%
Avg cost / sample$0.0011$0.0033
Avg speed / sample9.45s31.23s
By task
Object Detection (low)
42.0%
±1.1, Mean of 3 runs, range 40.9 to 43.1
$0.0014
59.3%
±0.7, Mean of 3 runs, range 58.5 to 60.0
$0.0028
Object Detection (high)
46.7%
±0.8, Mean of 3 runs, range 45.7 to 47.3
$0.0030
50.2%
±2.5, Mean of 3 runs, range 48.0 to 53.0
$0.023
Counting (low)
50.0%
±2.0, Mean of 3 runs, range 48.6 to 52.7
$0.0005
54.5%
±0.7, Mean of 3 runs, range 54.0 to 55.4
$0.0011
Counting (high)
59.0%
±5.4, Mean of 3 runs, range 52.7 to 63.5
$0.0016
63.1%
±2.0, Mean of 3 runs, range 60.8 to 64.9
$0.0095
Identification (low)
76.0%
±1.6, Mean of 3 runs, range 75.0 to 78.1
$0.0004
85.4%
±3.1, Mean of 3 runs, range 81.3 to 87.5
$0.0009
Identification (high)
78.1%
±4.7, Mean of 3 runs, range 71.9 to 81.3
$0.0012
85.4%
±1.6, Mean of 3 runs, range 84.4 to 87.5
$0.0044
OCR (low)
48.6%
$0.0012
46.1%
$0.0044
by category
Single value
37.0%
Transcription
68.2%
Structured JSON
67.7%
Text localization
18.7%
Single value
33.5%
Transcription
68.9%
Structured JSON
71.7%
Text localization
0.0%
OCR (high)
49.9%
$0.0034
54.5%
$0.021
by category
Single value
40.4%
Transcription
59.9%
Structured JSON
70.2%
Text localization
16.8%
Single value
38.3%
Transcription
75.3%
Structured JSON
76.4%
Text localization
28.8%
Reasoning (low)
35.1%
±2.6, Mean of 3 runs, range 32.5 to 37.8
$0.0005
38.9%
±0.3, Mean of 3 runs, range 38.4 to 39.1
$0.0013
Reasoning (high)
55.9%
±2.3, Mean of 3 runs, range 54.3 to 58.9
$0.0026
57.6%
±2.0, Mean of 3 runs, range 55.6 to 59.6
$0.013

MiMo V2.6 Pro vs Mistral Large 4: Overview

MiMo V2.6 Pro

MiMo V2.6 Pro is the flagship omni-modal foundation model in Xiaomi's MiMo V2.6 series, released as open weights alongside a Flash variant and a 9B distillation of Qwen3.5. It uses a sparse mixture-of-experts transformer with 1.02 trillion total parameters and roughly 42 billion activated per token, paired with a hybrid attention design that interleaves sliding-window and global attention layers to support a context window of about one million tokens. Dedicated encoders handle non-text inputs, including a vision encoder of roughly 681 million parameters and an audio tokenizer stack, so the model accepts text, images, video, and audio and returns text.

Post-training centers on large-scale reinforcement learning across thousands of interactive environments, combined with agentic grading, self-correction cold start, and a multi-prefix multi-teacher on-policy distillation stage that extends behavior to tasks that are hard to verify automatically.

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, Mistral Large 4 performed better. It scores higher on 4 of the five vision tasks and averages 56.8% (#46 of 68) against 50.3% (#59 of 68) for MiMo V2.6 Pro. These scores measure vision capabilities only; either model may perform differently on work outside this benchmark's tasks.

No. On the Vision Evals Object Detection benchmark at low effort, Mistral Large 4 leads with 59.3% against 42.0%. This is the widest gap between the two models across the benchmark's tasks.

MiMo V2.6 Pro is cheaper to run. Across Roboflow's Vision Evals task mix it averaged $0.0011 per sample against $0.0033. MiMo V2.6 Pro is priced at $0.43 per 1M input tokens and $0.87 per 1M output; Mistral Large 4 is priced at $0.70 per 1M input tokens and $2.10 per 1M output. Actual costs depend on your image sizes, prompts, and output length.

MiMo V2.6 Pro is faster. Across Roboflow's Vision Evals it averaged 9.5s 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 unlocks unlimited runs.