Coding
Like-for-like- Laguna M.1
- 64.8
- MiniMax M3
- 72.2
- Weighted basis
- 2 vs 2 rows
- Reading
- MiniMax M3 leads
Laguna M.1 vs MiniMax M3
Updated August 30, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner.
4 results are shared. Category rows based on different benchmark sets are marked directional and do not name a winner.
Which one for your work
Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.
Coding work
Code generation, repair, and software-engineering tasks
MiniMax M3
MiniMax M3 leads on the same 2 weighted benchmark rows.
Confidence: limited
Long documents
Prompts that approach the documented context limit
MiniMax M3
MiniMax M3 has the larger documented context window.
Confidence: documented
Show secondary and unsupported calls
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.
Confidence: limited
Chat turn cost
1K fresh input + 500 output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Repository review cost
50K fresh input + 3K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
1 Laguna M.1 only
4 Shared
18 MiniMax M3 only
- Shared results
- 4
- Laguna M.1 only
- 1
- MiniMax M3 only
- 18
- Like-for-like categories
- 1 / 8
1 category uses different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Category results, on a stated basis
Each row states whether both averages use the same weighted benchmark set. Directional and not-comparable rows remain visible, but they never receive a winner in this template.
Agentic
Directional only- Laguna M.1
- 45.8
- MiniMax M3
- 72.3
- Weighted basis
- 1 vs 3 rows
- Reading
- Directional only
Reasoning
Not comparable- Laguna M.1
- Not measured
- MiniMax M3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Knowledge
Not comparable- Laguna M.1
- Not measured
- MiniMax M3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Math
Not comparable- Laguna M.1
- Not measured
- MiniMax M3
- 85.7
- Weighted basis
- 0 vs 1 rows
- Reading
- Not comparable
Multilingual
Not comparable- Laguna M.1
- Not measured
- MiniMax M3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multimodal
Not comparable- Laguna M.1
- Not measured
- MiniMax M3
- 64.9
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Instruction following
Not comparable- Laguna M.1
- Not measured
- MiniMax M3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
| Category | Laguna M.1 | MiniMax M3 | Weighted basis | Reading |
|---|---|---|---|---|
| Coding | 64.8 | 72.2 | Like-for-like2 vs 2 rows | MiniMax M3 leads |
| Agentic | 45.8 | 72.3 | Directional only1 vs 3 rows | Directional only |
| Reasoning | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Knowledge | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Math | Not measured | 85.7 | Not comparable0 vs 1 rows | Not comparable |
| Multilingual | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Multimodal | Not measured | 64.9 | Not comparable0 vs 2 rows | Not comparable |
| Instruction following | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
Shape of the matched evidence
Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.
Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.
Terminal-Bench 2.0
Agentic
SWE-bench Pro
Coding
SWE-bench Verified
Coding
What each workload costs
Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.
Chat turn
1K fresh input + 500 output tokens
- Laguna M.1
- API rate not published
- Fits in one request
- MiniMax M3
- $0.0009
- Fits in one request
Laguna M.1 has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- Laguna M.1
- API rate not published
- Fits in one request
- MiniMax M3
- $0.0186
- Fits in one request
Laguna M.1 has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- Laguna M.1
- API rate not published
- Fits in one request
- Cached-input rate unavailable
- MiniMax M3
- $0.03
- Fits in one request
Laguna M.1 has no comparable published API token rate.
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationLaguna M.1MiniMax M3
Context window
Maximum documented context; output-token limits may be lower.
Laguna M.1
256K
MiniMax M3
1M
API model ID
Laguna M.1
Not sourced
MiniMax M3
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Laguna M.1
No comparable hosted API rate
MiniMax M3
$0.06 per 1M cached input tokens
Documented inputs
Laguna M.1
Not sourced
MiniMax M3
Not sourced
Documented outputs
Laguna M.1
Not sourced
MiniMax M3
Not sourced
Provider availability
Laguna M.1
Not sourced
MiniMax M3
Not sourced
Reasoning profile
Laguna M.1
Reasoning
MiniMax M3
Non-Reasoning
Weight access
Laguna M.1
Proprietary
MiniMax M3
Open Weight
License
Laguna M.1
Proprietary
MiniMax M3
Open Weight
Release date
Laguna M.1
2026-04-28
MiniMax M3
2026-06-01
If you already use one of these models
- Deployment change
- The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
- Quality signal
- At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner.
- Workload cost
- A complete comparable API-rate estimate is not available for both models.
- Context tradeoff
- MiniMax M3 has the larger documented window (1M).
Run the same representative tasks against both endpoints before changing production traffic.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence23 rows
Agentic
Terminal-Bench 2.0
Laguna M.145.8%SourceMiniMax M366%SourceMiniMax M3 leads this result
BrowseComp
Laguna M.1—MiniMax M383.5%SourceNot directly comparable
OSWorld-Verified
Laguna M.1—MiniMax M370.1%SourceNot directly comparable
MCP Atlas
Laguna M.1—MiniMax M374.2%SourceNot directly comparable
Claw-Eval
Laguna M.1—MiniMax M374.5%SourceNot directly comparable
BankerToolBench
Laguna M.1—MiniMax M376.1%SourceNot directly comparable
ResearchClawBench
Laguna M.1—MiniMax M319.8%SourceNot directly comparable
OSWorld 2.0
Laguna M.1—MiniMax M34.6%SourceNot directly comparable
Coding
SWE-bench Verified
Laguna M.174.6%SourceMiniMax M380.5%SourceMiniMax M3 leads this result
SWE Multilingual
Laguna M.163.1%SourceMiniMax M3—Not directly comparable
SWE-bench Pro
Laguna M.149.2%SourceMiniMax M359%SourceMiniMax M3 leads this result
Terminal-Bench 2.0
Laguna M.145.8%SourceMiniMax M366.0%SourceMiniMax M3 leads this result
NL2Repo
Laguna M.1—MiniMax M342.1%SourceNot directly comparable
VIBE V2
Laguna M.1—MiniMax M350.1%SourceNot directly comparable
SVG-Bench
Laguna M.1—MiniMax M363.7%SourceNot directly comparable
KernelBench Hard
Laguna M.1—MiniMax M328.8%SourceNot directly comparable
OpenHarmony Bench
Laguna M.1—MiniMax M348.4%SourceNot directly comparable
Math
USAMO 2026
Laguna M.1—MiniMax M385.7%SourceNot directly comparable
Multimodal
OfficeQA Pro
Laguna M.1—MiniMax M345.1%SourceNot directly comparable
OmniDocBench 1.5
Laguna M.1—MiniMax M391.6%SourceNot directly comparable
MMMU-Pro
Laguna M.1—MiniMax M378.1%SourceNot directly comparable
VideoMMMU
Laguna M.1—MiniMax M384.6%SourceNot directly comparable
Video-MME (with subtitle)
Laguna M.1—MiniMax M385.4%SourceNot directly comparable
Frequently asked questions
Which is better, Laguna M.1 or MiniMax M3?
At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner. The page therefore keeps the decision tied to the specific documented workload.
Which is better for coding, Laguna M.1 or MiniMax M3?
MiniMax M3 leads the like-for-like coding comparison across 2 shared weighted benchmark rows.
Which is better for agentic tasks, Laguna M.1 or MiniMax M3?
The current agentic tasks averages use different weighted benchmark sets, so BenchLM does not name a winner from them. Read the shared benchmark rows directly and test the models on the same task set.
Which costs less, Laguna M.1 or MiniMax M3?
Both models do not have comparable published API token rates, so this page does not name a universal price winner.
Which has the larger context window, Laguna M.1 or MiniMax M3?
MiniMax M3 has the larger documented context window: 1M, compared with 256K.
Related comparisons
Last updated August 30, 2026
Watch Laguna M.1 vs MiniMax M3
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.