Skip to main content
BenchLM
Advertise

Model comparison

Claude Opus 4.8 vs MiniMax M3

Data verified

Head-to-head evidence from 32 shared benchmark results across 7 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

77.8/100
Margin
8.0pts
← winning
MiniMax M3
MiniMax
69.8/100
3 category wins1 category wins

Verified leaderboard positions: Claude Opus 4.8 #1; MiniMax M3 #18

BenchAlign evidence: Claude Opus 4.8 supported; MiniMax M3 supported. Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. Claude Opus 4.8 and MiniMax M3 share 32 comparable benchmark results. 4 of 8 categories are comparable. 21 results are unique to Claude Opus 4.8; 13 to MiniMax M3.

Updated July 16, 2026
Shared results
32
Claude Opus 4.8 only
21
MiniMax M3 only
13
Comparable categories
4 / 8

Pick Claude Opus 4.8 if you want the stronger benchmark profile. MiniMax M3 only becomes the better choice if mathematics is the priority or you want the cheaper token bill.

Confidence note. This is a partial-evidence comparison with 32 shared benchmark results across 7 evidence categories; 4 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

Claude Opus 4.8 is clearly ahead on the provisional aggregate, 84 to 70. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Claude Opus 4.8's sharpest advantage is in multimodal & grounded, where it averages 77 against 64.9. The single biggest benchmark swing on the page is OfficeQA Pro, 66.2% to 45.1%. MiniMax M3 does hit back in mathematics, so the answer changes if that is the part of the workload you care about most.

Claude Opus 4.8 is also the more expensive model on tokens at $5.00 input / $25.00 output per 1M tokens, versus $0.30 input / $1.20 output per 1M tokens for MiniMax M3. That is roughly 20.8x on output cost alone. Claude Opus 4.8 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.

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 Claude Opus 4.8 and MiniMax M3
CategoryClaude Opus 4.8ΔMiniMax M3
MathClaude Opus 4.853.9Margin→ 31.8MiniMax M385.7
MultimodalClaude Opus 4.877.0Margin← 12.1MiniMax M364.9
CodingClaude Opus 4.881.1Margin← 8.9MiniMax M372.2
AgenticClaude Opus 4.880.3Margin← 8.0MiniMax M372.3
KnowledgeClaude Opus 4.862.7MarginNo overlapMiniMax M3Not measured

Decisive benchmark drivers

The largest measured benchmark gaps in this matchup, with exact reported values.

More
A · Claude Opus 4.8B · MiniMax M3

  1. OfficeQA Pro

    Multimodal
    Source ↗
    A 66.2%B 45.1%
    Winner: Claude Opus 4.8Δ 21.1
    OfficeQA Pro: Claude Opus 4.8 scored 66.2%; MiniMax M3 scored 45.1%. Claude Opus 4.8 wins this benchmark.

  2. OSWorld-Verified

    Agentic
    Source ↗
    A 83.4%B 70.1%
    Winner: Claude Opus 4.8Δ 13.3
    OSWorld-Verified: Claude Opus 4.8 scored 83.4%; MiniMax M3 scored 70.1%. Claude Opus 4.8 wins this benchmark.

  3. USAMO 2026

    Math
    Source ↗
    A 96.7%B 85.7%
    Winner: Claude Opus 4.8Δ 11
    USAMO 2026: Claude Opus 4.8 scored 96.7%; MiniMax M3 scored 85.7%. Claude Opus 4.8 wins this benchmark.

  4. SWE-bench Pro

    Coding
    Source ↗
    A 69.2%B 59%
    Winner: Claude Opus 4.8Δ 10.2
    SWE-bench Pro: Claude Opus 4.8 scored 69.2%; MiniMax M3 scored 59%. Claude Opus 4.8 wins this benchmark.

  5. Terminal-Bench 2.0

    Agentic
    Source ↗
    A 74.6%B 66%
    Winner: Claude Opus 4.8Δ 8.6
    Terminal-Bench 2.0: Claude Opus 4.8 scored 74.6%; MiniMax M3 scored 66%. Claude Opus 4.8 wins this benchmark.

Operational comparison

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

MetricClaude Opus 4.8MiniMax M3Comparison
Input / output priceUSD per 1M tokensClaude Opus 4.8$5 input / $25 outputMiniMax M3$0.3 input / $1.2 outputMiniMax M3 has the lower combined listed price.
Generation speedtokens per secondClaude Opus 4.8Not availableMiniMax M3Not availableA complete speed comparison is not available.
First-answer latencyseconds to first tokenClaude Opus 4.8Not availableMiniMax M3Not availableA complete latency comparison is not available.
Context windowmaximum listed tokensClaude Opus 4.81MMiniMax M31MListed context windows are equal.

Benchmark Deep Dive

AgenticClaude Opus 4.8 wins
BenchmarkClaude Opus 4.8MiniMax M3Result
Terminal-Bench 2.0Source 74.6%66%Claude Opus 4.8 leads
BrowseCompSource 84.3%83.5%Claude Opus 4.8 leads
DeepSearchQASource 93.1%—Not comparable
OSWorld-VerifiedSource 83.4%70.1%Claude Opus 4.8 leads
Finance Agent v2Source 53.9%—Not comparable
GDPval-AASource 16001395Claude Opus 4.8 leads
MCP AtlasSource 82.2%74.2%Claude Opus 4.8 leads
ToolathlonSource 59.9%—Not comparable
Gert LabsSource 72.97%—Not comparable
AA Agentic IndexSource 47.2%35.4%Claude Opus 4.8 leads
τ²-bench resultsSource 94.4%88.9%Claude Opus 4.8 leads
GDPval-AASource 55.0%44.7%Claude Opus 4.8 leads
ResearchClawBenchSource 21.1%19.8%Claude Opus 4.8 leads
OSWorld 2.0Source 20.6%4.6%Claude Opus 4.8 leads
AA BriefcaseSource 13541110Claude Opus 4.8 leads
AA AutomationBenchSource 48.5%—Not comparable
AA EnterpriseOps-GymSource 44.0%32.1%Claude Opus 4.8 leads
AA Harvey LABSource 7.5%6.7%Claude Opus 4.8 leads
AA Tau3 BankingSource 27.6%—Not comparable
Claw-EvalSource —74.5%Not comparable
GDPval rubricsSource —74.7%Not comparable
BankerToolBenchSource —76.1%Not comparable
CodingClaude Opus 4.8 wins
BenchmarkClaude Opus 4.8MiniMax M3Result
SWE-bench VerifiedSource 88.6%80.5%Claude Opus 4.8 leads
SWE-bench ProSource 69.2%59%Claude Opus 4.8 leads
SWE MultilingualSource 84.4%—Not comparable
SWE MultimodalSource 38.4%—Not comparable
Terminal-Bench 2.0Source 74.6%66.0%Claude Opus 4.8 leads
cursorBench31Source 58.4%—Not comparable
cursorBench32Source 62.3%—Not comparable
AA Coding IndexSource 74.3%58.6%Claude Opus 4.8 leads
Terminal-Bench HardSource 58.3%42.4%Claude Opus 4.8 leads
AA-SciCodeSource 53.5%45.4%Claude Opus 4.8 leads
FrontierCodeSource 46.5%—Not comparable
AA Terminal-Bench 2.1Source 84.6%65.2%Claude Opus 4.8 leads
NL2RepoSource —42.1%Not comparable
VIBE V2Source —50.1%Not comparable
SVG-BenchSource —63.7%Not comparable
KernelBench HardSource —28.8%Not comparable
Reasoning
BenchmarkClaude Opus 4.8MiniMax M3Result
AA-LCRSource 67.7%74.0%MiniMax M3 leads
CritPtSource 20.9%3.7%Claude Opus 4.8 leads
Knowledge
BenchmarkClaude Opus 4.8MiniMax M3Result
GPQASource 93.6%—Not comparable
GPQA-DSource 93.6%—Not comparable
HLESource 57.9%—Not comparable
HLE w/o toolsSource 49.8%—Not comparable
Artificial Analysis Intelligence IndexSource 55.7%44.4%Claude Opus 4.8 leads
AA-GPQA DiamondSource 92.0%92.9%MiniMax M3 leads
AA-HLESource 45.7%37.1%Claude Opus 4.8 leads
AA-Omniscience IndexSource 27.4%1.4%Claude Opus 4.8 leads
AA-Omniscience AccuracySource 46.6%15.0%Claude Opus 4.8 leads
AA-Omniscience Hallucination RateSource 35.9%16.1%MiniMax M3 leads
AA Openness IndexSource —33.3%Not comparable
MathMiniMax M3 wins
BenchmarkClaude Opus 4.8MiniMax M3Result
USAMO 2026Source 96.7%85.7%Claude Opus 4.8 leads
FrontierMath v2 (Tiers 1-3)Source 47.241%—Not comparable
FrontierMath v2 (Tier 4)Source 31.250%—Not comparable
Multilingual
BenchmarkClaude Opus 4.8MiniMax M3Result
INCLUDESource 87.6%—Not comparable
MultimodalClaude Opus 4.8 wins
BenchmarkClaude Opus 4.8MiniMax M3Result
OfficeQA ProSource 66.2%45.1%Claude Opus 4.8 leads
ScreenSpot ProSource 87.9%—Not comparable
CharXivSource 89.9%—Not comparable
CharXiv w/o toolsSource 80.5%—Not comparable
Design Arena WebsiteSource 12741294MiniMax M3 leads
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
AA-MMMU-ProSource —78.6%Not comparable
Inst. Following
BenchmarkClaude Opus 4.8MiniMax M3Result
AA-IFBenchSource 62.2%82.9%MiniMax M3 leads
Frequently Asked Questions (5)

Which is better, Claude Opus 4.8 or MiniMax M3?

Claude Opus 4.8 is ahead on BenchLM's provisional leaderboard, 84 to 70. The biggest single separator in this matchup is OfficeQA Pro, where the scores are 66.2% and 45.1%.

Which is better for coding, Claude Opus 4.8 or MiniMax M3?

Claude Opus 4.8 has the edge for coding in this comparison, averaging 81.1 versus 72.2. Inside this category, AA Terminal-Bench 2.1 is the benchmark that creates the most daylight between them.

Which is better for math, Claude Opus 4.8 or MiniMax M3?

MiniMax M3 has the edge for math in this comparison, averaging 85.7 versus 53.9. Inside this category, USAMO 2026 is the benchmark that creates the most daylight between them.

Which is better for agentic tasks, Claude Opus 4.8 or MiniMax M3?

Claude Opus 4.8 has the edge for agentic tasks in this comparison, averaging 80.3 versus 72.3. Inside this category, AA Briefcase is the benchmark that creates the most daylight between them.

Which is better for multimodal and grounded tasks, Claude Opus 4.8 or MiniMax M3?

Claude Opus 4.8 has the edge for multimodal and grounded tasks in this comparison, averaging 77 versus 64.9. Inside this category, OfficeQA Pro is the benchmark that creates the most daylight between them.

Related Comparisons

Explore More

Anthropic Compare Pricing Methodology Find Your Best LLM Overall Rankings
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.