Skip to main content

Model comparison

LFM2.5-8B-A1B vs MiniMax M3

Data verified

Head-to-head evidence from 12 shared benchmark results across 5 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

41.25/100
Margin
28.5pts
winning →
MiniMax
69.8/100
0 category wins1 category wins

Verified leaderboard positions: LFM2.5-8B-A1B unranked; MiniMax M3 #18

BenchAlign evidence: LFM2.5-8B-A1B estimated; MiniMax M3 supported. Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. LFM2.5-8B-A1B and MiniMax M3 share 12 comparable benchmark results. 1 of 8 categories are comparable. 6 results are unique to LFM2.5-8B-A1B; 33 to MiniMax M3.

Updated July 16, 2026
Shared results
12
LFM2.5-8B-A1B only
6
MiniMax M3 only
33
Comparable categories
1 / 8

Pick MiniMax M3 if you want the stronger benchmark profile. LFM2.5-8B-A1B only becomes the better choice if you want the cheaper token bill or you want the stronger reasoning-first profile.

Confidence note. This is a partial-evidence comparison with 12 shared benchmark results across 5 evidence categories; 1 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

MiniMax M3 is clearly ahead on the provisional aggregate, 70 to 37. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

MiniMax M3's sharpest advantage is in mathematics, where it averages 85.7 against 50.

MiniMax M3 is also the more expensive model on tokens at $0.30 input / $1.20 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for LFM2.5-8B-A1B. That is roughly Infinityx on output cost alone. LFM2.5-8B-A1B is the reasoning model in the pair, while MiniMax M3 is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use. MiniMax M3 gives you the larger context window at 1M, compared with 128K for LFM2.5-8B-A1B.

Category breakdown

Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.

Category scores and score margins for LFM2.5-8B-A1B and MiniMax M3
CategoryLFM2.5-8B-A1BΔMiniMax M3
MathLFM2.5-8B-A1B50.0Margin 35.7MiniMax M385.7
AgenticLFM2.5-8B-A1BNot measuredMarginNo overlapMiniMax M372.3
CodingLFM2.5-8B-A1BNot measuredMarginNo overlapMiniMax M372.2
MultimodalLFM2.5-8B-A1BNot measuredMarginNo overlapMiniMax M364.9
Inst. FollowingLFM2.5-8B-A1B68.8MarginNo overlapMiniMax M3Not measured

Operational comparison

Runtime and commercial metrics are compared only when both models have a complete sourced value.

MetricLFM2.5-8B-A1BMiniMax M3Comparison
Input / output priceUSD per 1M tokensLFM2.5-8B-A1B$0 input / $0 outputMiniMax M3$0.3 input / $1.2 outputLFM2.5-8B-A1B has the lower combined listed price.
Generation speedtokens per secondLFM2.5-8B-A1BNot availableMiniMax M3Not availableA complete speed comparison is not available.
First-answer latencyseconds to first tokenLFM2.5-8B-A1BNot availableMiniMax M3Not availableA complete latency comparison is not available.
Context windowmaximum listed tokensLFM2.5-8B-A1B128KMiniMax M31MMiniMax M3 lists the larger context window.

Benchmark Deep Dive

Agentic
BenchmarkLFM2.5-8B-A1BMiniMax M3Result
BFCL v4Source 49.7%Not comparable
τ²-bench resultsSource 16.1%88.9%MiniMax M3 leads
Terminal-Bench 2.0Source 66%Not comparable
BrowseCompSource 83.5%Not comparable
OSWorld-VerifiedSource 70.1%Not comparable
MCP AtlasSource 74.2%Not comparable
Claw-EvalSource 74.5%Not comparable
AA Agentic IndexSource 35.4%Not comparable
GDPval-AASource 44.7%Not comparable
GDPval-AASource 1395Not comparable
GDPval rubricsSource 74.7%Not comparable
BankerToolBenchSource 76.1%Not comparable
ResearchClawBenchSource 19.8%Not comparable
OSWorld 2.0Source 4.6%Not comparable
AA BriefcaseSource 1110Not comparable
AA EnterpriseOps-GymSource 32.1%Not comparable
AA Harvey LABSource 6.7%Not comparable
Coding
BenchmarkLFM2.5-8B-A1BMiniMax M3Result
Terminal-Bench HardSource 4.5%42.4%MiniMax M3 leads
AA-SciCodeSource 7.8%45.4%MiniMax M3 leads
SWE-bench VerifiedSource 80.5%Not comparable
SWE-bench ProSource 59%Not comparable
Terminal-Bench 2.0Source 66.0%Not comparable
NL2RepoSource 42.1%Not comparable
AA Coding IndexSource 58.6%Not comparable
VIBE V2Source 50.1%Not comparable
SVG-BenchSource 63.7%Not comparable
KernelBench HardSource 28.8%Not comparable
AA Terminal-Bench 2.1Source 65.2%Not comparable
Reasoning
BenchmarkLFM2.5-8B-A1BMiniMax M3Result
AA-LCRSource 0.0%74.0%MiniMax M3 leads
CritPtSource 0.0%3.7%MiniMax M3 leads
Knowledge
BenchmarkLFM2.5-8B-A1BMiniMax M3Result
AA-GPQA DiamondSource 51.3%92.9%MiniMax M3 leads
AA-HLESource 6.9%37.1%MiniMax M3 leads
AA-Omniscience IndexSource -33.3%1.4%MiniMax M3 leads
AA-Omniscience AccuracySource 9.4%15.0%MiniMax M3 leads
AA-Omniscience Hallucination RateSource 47.0%16.1%MiniMax M3 leads
Artificial Analysis Intelligence IndexSource 8.3%44.4%MiniMax M3 leads
AA Openness IndexSource 33.3%Not comparable
MathMiniMax M3 wins
BenchmarkLFM2.5-8B-A1BMiniMax M3Result
MATH-500Source 88.8%Not comparable
AIME 2025Source 42.5%Not comparable
AIME26Source 50.0%Not comparable
USAMO 2026Source 85.7%Not comparable
Multimodal
BenchmarkLFM2.5-8B-A1BMiniMax M3Result
OfficeQA ProSource 45.1%Not comparable
OmniDocBench 1.5Source 91.6%Not comparable
MMMU-ProSource 78.1%Not comparable
VideoMMMUSource 84.6%Not comparable
Video-MME (with subtitle)Source 85.4%Not comparable
Design Arena WebsiteSource 1294Not comparable
AA-MMMU-ProSource 78.6%Not comparable
Inst. Following
BenchmarkLFM2.5-8B-A1BMiniMax M3Result
IFEvalSource 91.8%Not comparable
IFBenchSource 56.5%Not comparable
AA-IFBenchSource 55.6%82.9%MiniMax M3 leads
Frequently Asked Questions (2)

Which is better, LFM2.5-8B-A1B or MiniMax M3?

MiniMax M3 is ahead on BenchLM's provisional leaderboard, 70 to 37.

Which is better for math, LFM2.5-8B-A1B or MiniMax M3?

MiniMax M3 has the edge for math in this comparison, averaging 85.7 versus 50. LFM2.5-8B-A1B stays close enough that the answer can still flip depending on your workload.

Related Comparisons

Last updated: July 16, 2026

The AI models change fast. We track them for you.

A weekly brief for engineers and researchers covering new models, ranking shifts, and pricing changes.

Free. No spam. Unsubscribe anytime.