EyesInAI
Switchyard
Groundyard
Connect
Benchmarks
Models
Learn
News
Home
Switchyard
Groundyard
Connect
Which model for your task?
Leaderboard
Deep Suite
Coding Models
Bake-offs
Measured
Compare
Trends
Try Models
Catalog
Providers
By Origin
By Layer
Hardware
Open Source
License Tiers
What we’re building
Explainers
Automation & Agents
Cut Costs
Case Studies
Papers
Notes
Principles
Methodology
Jargon
Help
Contact
Latest
Governance & Access
Pricing & Billing
Data Centers & Power
Articles
Investigated & Declined
Trending models
Status
Archive
Benchmark your use case
Loading benchmark data…
EyesInAI
·
Loading live benchmark data
Home
Benchmark
Settings
Loading models…
EyesInAI
·
Loading models…
All models
openrouter
minimax/minimax-m3
Last tested Jul 23, 2026
Overall pass
46%
Avg latency
2247 ms
Context
1000k
Tools
Yes
Input $/1M
$0.30
Output $/1M
$1.20
Tests run
13
Passed
6/13
AI summary
Hardware to run it
Hosted:
$0.3 in / $1.2 out
per 1M
Self-host:
API only
Proprietary / weights not released — hosted API only.
Test results
⚡
Ping
Latency & availability — single-word reply
939 ms
🧮
Reasoning
Basic math reasoning — show work, give answer
1699 ms
{ }
JSON Output
Structured output compliance — valid JSON with required keys
6970 ms
💻
Code Gen
Python function generation with docstring
fail
🚀
Throughput
Token generation speed — 500-token long-form response
368 ms
🔍
Context Recall
Retrieval from in-context data — 20-item list Q&A
Compare with another model
See the leaderboard
2947 ms
📝
Summarization
Real-world: distill a news article into 3 bullet points
fail
🏷️
Classification
Real-world: sentiment + category from customer review
fail
📄
Data Extraction
Real-world: pull structured fields from an invoice
2352 ms
📋
Instruction Follow
Real-world: multi-rule compliance (5 sentences, no "the", etc.)
fail
✅
Format Compliance
IFEval-style: 4 bullets, keyword inclusion/exclusion, no preamble
fail
🪡
Long-Context Needle
Find a 6-digit code buried in ~3.5k tokens of filler text
fail
🧠
Multi-Step Logic
BBH-lite: boolean expression + web of lies + object counting
fail