MiniMax H3
MiniMaxMiniMax
Release Date
2026-07
Parameters
—
Context Length
205K
Modalities
text
Capability Radar
60
general
60
coding
60
reasoning
51
scienceest.
70
agents
0
multimodal
Science uses a reasoning proxy when dedicated science benchmarks are unavailable.
Rankings
| Domain | #Rank | Score | Source |
|---|---|---|---|
| Agentic Capability | 99 | 51.0 | LS |
| Image-to-Video | 2 | 96.0 | AA |
| Text-to-Video | 2 | 97.0 | AA |
Benchmark Scores (LLM Stats)
Agents
BrowseComp
62.0%SR
Terminal-Bench
47.9%SR
Toolathlon
43.5%SR
Biology
GPQA
81.0%SR
SciCode
39.0%SR
Code
VIBE Web
91.5%SR
VIBE Android
89.7%SR
VIBE
88.6%SR
VIBE iOS
88.0%SR
VIBE Simulation
87.1%SR
VIBE Backend
86.7%SR
LiveCodeBench
78.0%SR
SWE-bench Multilingual
72.5%SR
SWT-Bench
69.3%SR
SWE-Bench Verified
67.0%SR
Multi-SWE-Bench
49.4%SR
OctoCodingBench
26.1%SR
SWE-Review
8.9%SR
SWE-Perf
3.1%SR
Communication
Tau2 Telecom
87.0%SR
Finance
MMLU-Pro
88.0%SR
General
IFBench
70.0%SR
Long Context
AA-LCR
62.0%SR
Math
AIME 2025
81.0%SR
Humanity's Last Exam
22.0%SR
AA Evaluation Indices
No AA evaluation data available
LLM Stats Category Scores
Language90
Legal90
Finance90
Healthcare90
Communication90
Instruction Following70
Frontend Development70
Tool Calling70
Long Context60
Math60
Physics60
Reasoning60
Search60
General60
Biology60
Chemistry60
Code60
Agents50
Vision20
Pricing
Input Price$0.3 / 1M tokens
Output Price$1.2 / 1M tokens
Blended Price (3:1)$0.525 / 1M tokens
Speed
No speed data available
Provider Price Ranking
Provider Price Ranking
1 providers
ProviderInputOutput
1MiniMaxPRIMARY
$0.3
$1.2
Compare pricing across different API providers for this model.