MiniMax M3 vs Step 3.5 Flash

Updated October 23, 2026. Rank says MiniMax M3 is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty.

Decision reading

MiniMax M3 has the higher public point estimate, 54.82 versus 42.03. Their conditional score ranges overlap. These ranges do not establish rank confidence. 1 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.

Model A

MiniMax M3

MiniMax

54.82/100

Supported · Public rank #70

90% interval 45.4–64.2

Model B

Step 3.5 Flash

StepFun

42.03/100

Estimated · Public rank #128

Conditional range 32.3–51.8

Shared results: 1
MiniMax M3 only: 27
Step 3.5 Flash only: 2
Like-for-like categories: 0 / 8

Supported: MiniMax M3 · Estimated: Step 3.5 Flash. Conditional ranges do not establish rank confidence.How the comparison works

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.

Long documents
Prompts that approach the documented context limit
MiniMax M3
MiniMax M3 has the larger documented context window.
Confidence: documented
Chat turn cost
1K fresh input + 500 output tokens
Step 3.5 Flash
Step 3.5 Flash has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Step 3.5 Flash
Step 3.5 Flash has the lower estimated token cost for this stated workload. Step 3.5 Flash has no published cached-input rate, so cached tokens use its listed input rate.
Confidence: rate-fallback

Show secondary and unsupported calls

Repository review cost
50K fresh input + 3K output tokens
Step 3.5 Flash
Step 3.5 Flash has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
Step 3.5 Flash is not ranked on the public lane for coding, so no winner is named for coding.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
Step 3.5 Flash is not ranked on the public lane for agentic, so no winner is named for agentic.
Confidence: limited

Which one for a specific job

Choose a job from the LLM Selector's task catalog to see the category row it rests on, under the same basis rules as the table below. A directional row stays directional; choosing a job never creates a winner.

Pick the job you would use them for

The same task catalog as the LLM Selector. Each job names the evidence surface it rests on; nothing here adds a new score.

37.0MiniMax M3—Step 3.5 Flash

Not comparable · BenchAlign v5.8

The coding row is not comparable on the public lane: at least one model is not measured or not ranked there.

Coding scores combine specific tasks and setups. Match the editor, harness, and effort to your workflow.

Open the coding leaderboard Shortlist models for this job

Same basis rules as the category table below

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.

SWE-bench VerifiedCoding
Normalized gap 6.1
MiniMax M3:80.5%
Step 3.5 Flash:74.4%
MiniMax M3 source Step 3.5 Flash source

Bars run 0–100 on each benchmark’s normalized display scale

Category results, on a stated basis

Each row shows the public-lane category score for both models: the BenchAlign v5.8 lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.

Agentic

Not comparable

MiniMax M3: 39.2; Supported · #71/132
Step 3.5 Flash: Not ranked
Basis: BenchAlign v5.8 lane · 10 vs 0 public rows
Reading: Not comparable

Coding

Not comparable

MiniMax M3: 37.0; Supported · #75/153
Step 3.5 Flash: Not ranked
Basis: BenchAlign v5.8 lane · 10 vs 2 public rows
Reading: Not comparable

Reasoning

Not comparable

MiniMax M3: 82.0; Unranked · 2 rankable rows
Step 3.5 Flash: Not ranked
Basis: Provisional lane · 0 vs 0 weighted rows
Reading: Not comparable

Multimodal

Not comparable

MiniMax M3: 59.2; #46/56
Step 3.5 Flash: Not ranked
Basis: Provisional lane · 2 vs 0 weighted rows
Reading: Not comparable

Knowledge

Not comparable

MiniMax M3: 51.0; Supported · #76/187
Step 3.5 Flash: Not ranked
Basis: BenchAlign v5.8 lane · 2 vs 0 public rows
Reading: Not comparable

Multilingual

Not comparable

MiniMax M3: Not ranked
Step 3.5 Flash: Not ranked
Basis: Provisional lane · 0 vs 0 weighted rows
Reading: Not comparable

Instruction following

Not comparable

MiniMax M3: 93.8; #6/153
Step 3.5 Flash: Not ranked
Basis: Provisional lane · 0 vs 0 weighted rows
Reading: Not comparable

Math

Not comparable

MiniMax M3: Not ranked
Step 3.5 Flash: Not ranked
Basis: Provisional lane · 1 vs 0 weighted rows
Reading: Not comparable

Category scores with the server-provided evidence basis for MiniMax M3 and Step 3.5 Flash
Category	MiniMax M3	Step 3.5 Flash	Basis	Reading
Agentic	39.2Supported · #71/132	Not ranked	Not comparableBenchAlign v5.8 lane · 10 vs 0 public rows	Not comparable
Coding	37.0Supported · #75/153	Not ranked	Not comparableBenchAlign v5.8 lane · 10 vs 2 public rows	Not comparable
Reasoning	82.0Unranked · 2 rankable rows	Not ranked	Not comparableProvisional lane · 0 vs 0 weighted rows	Not comparable
Multimodal	59.2#46/56	Not ranked	Not comparableProvisional lane · 2 vs 0 weighted rows	Not comparable
Knowledge	51.0Supported · #76/187	Not ranked	Not comparableBenchAlign v5.8 lane · 2 vs 0 public rows	Not comparable
Multilingual	Not ranked	Not ranked	Not comparableProvisional lane · 0 vs 0 weighted rows	Not comparable
Instruction following	93.8#6/153	Not ranked	Not comparableProvisional lane · 0 vs 0 weighted rows	Not comparable
Math	Not ranked	Not ranked	Not comparableProvisional lane · 1 vs 0 weighted rows	Not comparable

Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.8) differ from the provisional-lane categories. Unranked scores sit on the provisional lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.

Supported evidence per lane · bars run 0–100Methodology

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

MiniMax M3: $0.0009; Fits in one request
Step 3.5 Flash: $0.00025; Fits in one request

Step 3.5 Flash has the lower modeled cost

Costs use the listed standard API rates.

Repository review

50K fresh input + 3K output tokens

MiniMax M3: $0.0186; Fits in one request
Step 3.5 Flash: $0.0059; Fits in one request

Step 3.5 Flash has the lower modeled cost

Costs use the listed standard API rates.

Cache-heavy agent loop

200K cached + 20K fresh input + 10K output tokens

MiniMax M3: $0.03; Fits in one request
Step 3.5 Flash: $0.025; Fits in one request; Cached input priced at the published list-input rate

Step 3.5 Flash has the lower modeled cost

Step 3.5 Flash has no published cached-input rate, so cached tokens use its listed input rate.

Cached input falls back to the list input rate only where a cached rate is unpublished

Specification differences

Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.

SpecificationMiniMax M3Step 3.5 Flash

Context window

Maximum documented context; output-token limits may be lower.

MiniMax M3

Step 3.5 Flash

256K

API model ID

MiniMax M3

Not sourced

Step 3.5 Flash

Not sourced

Cached-input rate

A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.

MiniMax M3

$0.06 per 1M cached input tokens

Step 3.5 Flash

Not published

Documented inputs

MiniMax M3

Not sourced

Step 3.5 Flash

Not sourced

Documented outputs

MiniMax M3

Not sourced

Step 3.5 Flash

Not sourced

Provider availability

MiniMax M3

Not sourced

Step 3.5 Flash

Not sourced

Reasoning profile

MiniMax M3

Non-Reasoning

Step 3.5 Flash

Non-Reasoning

Weight access

MiniMax M3

Open Weight

Step 3.5 Flash

Open Weight

License

MiniMax M3

Open Weight

Step 3.5 Flash

Open Weight

Release date

MiniMax M3

2026-06-01

Step 3.5 Flash

2026-02-01

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: MiniMax M3 has the higher public point estimate, 54.82 versus 42.03. Their conditional score ranges overlap. These ranges do not establish rank confidence.
Workload cost: Repository review: $0.0186 vs $0.0059. Cache-heavy agent loop: $0.03 vs $0.025.
Context tradeoff: MiniMax M3 has the larger documented window (1M).

Run the same representative tasks against both endpoints before changing production traffic.

Questions

Which is better, MiniMax M3 or Step 3.5 Flash?

MiniMax M3 has the higher public point estimate, 54.82 versus 42.03. Their conditional score ranges overlap. These ranges do not establish rank confidence. The page therefore keeps the decision tied to the specific documented workload.

Which is better for coding, MiniMax M3 or Step 3.5 Flash?

Step 3.5 Flash is not ranked on the public lane for coding, so no winner is named for coding.

Which is better for agentic tasks, MiniMax M3 or Step 3.5 Flash?

Step 3.5 Flash is not ranked on the public lane for agentic tasks, so no winner is named for agentic tasks.

Which costs less, MiniMax M3 or Step 3.5 Flash?

For the stated presets, chat costs $0.0009 on MiniMax M3 and $0.00025 on Step 3.5 Flash; repository review costs $0.0186 and $0.0059; the cache-heavy agent loop costs $0.03 and $0.025. Step 3.5 Flash has no published cached-input rate, so cached tokens use its listed input rate.

Which has the larger context window, MiniMax M3 or Step 3.5 Flash?

MiniMax M3 has the larger documented context window: 1M, compared with 256K.

Self-host vs API cost

Estimates at 50,000 req/day · 1000 tokens/req average.

MiniMax M3

API / mo$1,125

Self-host / mo$13,981

Break-even818M/day

Step 3.5 Flash

API / mo$300

Self-host / moNot listed

Break-even—

Proprietary model — self-hosting not applicable.

Model the full break-even

Benchmark evidence

The full public result ledger is available for audit without forcing a wide desktop table onto a phone.

Browse raw public benchmark evidence30 rows

Agentic

Terminal-Bench 2.1
MiniMax M366.0%
Source
Step 3.5 Flash—
Not directly comparable
BrowseComp
MiniMax M383.5%
Source
Step 3.5 Flash—
Not directly comparable
OSWorld-Verified
MiniMax M370.1%
Source
Step 3.5 Flash—
Not directly comparable
MCP Atlas
MiniMax M374.2%
Source
Step 3.5 Flash—
Not directly comparable
Claw-Eval
MiniMax M374.5%
Source
Step 3.5 Flash—
Not directly comparable
BankerToolBench
MiniMax M376.1%
Source
Step 3.5 Flash—
Not directly comparable
ResearchClawBench
MiniMax M319.8%
Source
Step 3.5 Flash—
Not directly comparable
OSWorld 2.0
MiniMax M34.6%
Source
Step 3.5 Flash—
Not directly comparable
Terminal-Bench 2.1 (Vals)
MiniMax M353.6%
Source
Step 3.5 Flash—
Not directly comparable
MazeBench
MiniMax M30%
Source
Step 3.5 Flash—
Not directly comparable

Coding

SWE-bench Verified
MiniMax M380.5%
Source
Step 3.5 Flash74.4%
Source
MiniMax M3 leads this result
SWE-bench Pro
MiniMax M359%
Source
Step 3.5 Flash—
Not directly comparable
Terminal-Bench 2.1
MiniMax M366.0%
Source
Step 3.5 Flash—
Not directly comparable
NL2Repo
MiniMax M342.1%
Source
Step 3.5 Flash—
Not directly comparable
VIBE V2
MiniMax M350.1%
Source
Step 3.5 Flash—
Not directly comparable
SVG-Bench
MiniMax M363.7%
Source
Step 3.5 Flash—
Not directly comparable
KernelBench Hard
MiniMax M328.8%
Source
Step 3.5 Flash—
Not directly comparable
OpenHarmony Bench
MiniMax M348.4%
Source
Step 3.5 Flash—
Not directly comparable
LiveCodeBench (Vals)
MiniMax M382.2%
Source
Step 3.5 Flash—
Not directly comparable
SWE-bench (Vals)
MiniMax M375.0%
Source
Step 3.5 Flash—
Not directly comparable
LiveCodeBench v6
MiniMax M3—
Step 3.5 Flash86.4%
Source
Not directly comparable

Multimodal

OfficeQA Pro
MiniMax M345.1%
Source
Step 3.5 Flash—
Not directly comparable
OmniDocBench 1.5
MiniMax M391.6%
Source
Step 3.5 Flash—
Not directly comparable
MMMU-Pro
MiniMax M378.1%
Source
Step 3.5 Flash—
Not directly comparable
VideoMMMU
MiniMax M384.6%
Source
Step 3.5 Flash—
Not directly comparable
Video-MME (with subtitle)
MiniMax M385.4%
Source
Step 3.5 Flash—
Not directly comparable

Knowledge

GPQA Diamond (Vals)
MiniMax M392.7%
Source
Step 3.5 Flash—
Not directly comparable
MMLU-Pro (Vals)
MiniMax M384.2%
Source
Step 3.5 Flash—
Not directly comparable

Math

USAMO 2026
MiniMax M385.7%
Source
Step 3.5 Flash—
Not directly comparable
AIME 2025
MiniMax M3—
Step 3.5 Flash97.3%
Source
Not directly comparable

30 public results · 1 shared

Related comparisons

Compare API pricing Read the methodology Open the model selector

Watch MiniMax M3 vs Step 3.5 Flash

One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.

Read a sample issue

Join 5,600+ readers.

Last updated October 23, 2026

MiniMax M3 vs Step 3.5 Flash

Decision reading

Which one for your work

Long documents

Chat turn cost

Cache-heavy agent loop cost

Repository review cost

Coding work

Agentic work

Which one for a specific job

What is actually comparable

Shape of the matched evidence

Category results, on a stated basis

Agentic

Coding

Reasoning

Multimodal

Knowledge

Multilingual

Instruction following

Math

What each workload costs

Chat turn

Repository review

Cache-heavy agent loop

Specification differences

Context window

API model ID

Cached-input rate

Documented inputs

Documented outputs

Provider availability

Reasoning profile

Weight access

License

Release date

If you already use one of these models

Questions

Which is better, MiniMax M3 or Step 3.5 Flash?

Which is better for coding, MiniMax M3 or Step 3.5 Flash?

Which is better for agentic tasks, MiniMax M3 or Step 3.5 Flash?

Which costs less, MiniMax M3 or Step 3.5 Flash?

Which has the larger context window, MiniMax M3 or Step 3.5 Flash?

Self-host vs API cost

Benchmark evidence

Agentic

Coding

Multimodal

Knowledge

Math

Related comparisons

Watch MiniMax M3 vs Step 3.5 Flash