Skip to main content
Radar

Five or fewer confirmed AI changes, with original sources, on mornings when something changed.A free source-linked morning brief.

See the free Radar Brief
BenchLM
Celeris logo
Model A
Celeris-1

Celeris

—

Evidence status unavailable

90% interval unavailable

Celeris-1 vs MiniMax M3

Updated September 4, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.

MiniMax logo
Model B
MiniMax M3

MiniMax

63.91/100

Supported · Public rank #51

90% interval 56.5–71.3

Decision reading

The public evidence has no benchmark result shared by both models, so it does not support a quality verdict. Use the documented cost, context, and runtime rows instead.

0 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.

Share or export

Share on XLinkedInSocial cardCSVJSON

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

Show secondary and unsupported calls

  • Coding work

    Code generation, repair, and software-engineering tasks

    Not enough matched evidence

    Celeris-1 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

    Celeris-1 is not ranked on the public lane for agentic, so no winner is named for agentic.

    Confidence: limited

  • Chat turn cost

    1K fresh input + 500 output tokens

    Not enough matched evidence

    The page does not recommend a cost winner because at least one model cannot fit the stated workload in one request. Celeris-1 does not fit this workload in one request.

    Confidence: listed-rates

  • Cache-heavy agent loop cost

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

    Not enough matched evidence

    The page does not recommend a cost winner because at least one model cannot fit the stated workload in one request. Celeris-1 does not fit this workload in one request. Celeris-1 has no published cached-input rate, so cached tokens use its listed input rate.

    Confidence: rate-fallback

  • Repository review cost

    50K fresh input + 3K output tokens

    Not enough matched evidence

    The page does not recommend a cost winner because at least one model cannot fit the stated workload in one request. Celeris-1 does not fit this workload in one request.

    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.

Shared results
0
Celeris-1 only
4
MiniMax M3 only
27
Like-for-like categories
0 / 8

Category results, on a stated basis

Each row shows the public-lane category score for both models: the BenchAlign 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
Celeris-1
Not ranked
MiniMax M3
43.2
Supported · #106/151
Basis
BenchAlign lane · 0 vs 9 public rows
Reading
Not comparable

Coding

Not comparable
Celeris-1
Not ranked
MiniMax M3
50.6
Estimated · #68/183
Basis
BenchAlign lane · 0 vs 10 public rows
Reading
Not comparable

Reasoning

Not comparable
Celeris-1
47.8
Unranked · 3 rankable rows
MiniMax M3
78.5
#5/22
Basis
Provisional lane · 0 vs 0 weighted rows
Reading
Not comparable

Knowledge

Not comparable
Celeris-1
Not ranked
MiniMax M3
53.9
Supported · #65/181
Basis
BenchAlign lane · 1 vs 2 public rows
Reading
Not comparable

Math

Not comparable
Celeris-1
Not ranked
MiniMax M3
Not ranked
Basis
Provisional lane · 0 vs 1 weighted rows
Reading
Not comparable

Multilingual

Not comparable
Celeris-1
Not ranked
MiniMax M3
Not ranked
Basis
Provisional lane · 0 vs 0 weighted rows
Reading
Not comparable

Multimodal

Not comparable
Celeris-1
Not ranked
MiniMax M3
52.0
#34/48
Basis
Provisional lane · 0 vs 2 weighted rows
Reading
Not comparable

Instruction following

Not comparable
Celeris-1
Not ranked
MiniMax M3
93.5
#4/120
Basis
Provisional lane · 0 vs 0 weighted rows
Reading
Not comparable

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

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.

A shared-evidence shape is not available.

BenchLM does not draw a radar or infer missing axes when the matched evidence is too sparse.

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

Celeris-1
$0.00055
Does not fit in one request
MiniMax M3
$0.0009
Fits in one request

Celeris-1 does not fit this workload in one request.

Repository review

50K fresh input + 3K output tokens

Celeris-1
$0.0121
Does not fit in one request
MiniMax M3
$0.0186
Fits in one request

Celeris-1 does not fit this workload in one request.

Cache-heavy agent loop

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

Celeris-1
$0.051
Does not fit in one request
Cached input priced at the published list-input rate
MiniMax M3
$0.03
Fits in one request

Celeris-1 does not fit this workload in one request. Celeris-1 has no published cached-input rate, so cached tokens use its listed input rate.

Specification differences

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

Context window

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

Celeris-1

131,072 tokens

Celeris-1 model guide

MiniMax M3

1M

Cached-input rate

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

Celeris-1

Not published

Celeris API pricing

MiniMax M3

$0.06 per 1M cached input tokens

Provider availability

Celeris-1

Generally Available · Celeris OpenAI-compatible API (United States)

Celeris availability

MiniMax M3

Not sourced

Reasoning profile

Celeris-1

Non-Reasoning

MiniMax M3

Non-Reasoning

Weight access

Celeris-1

Proprietary

MiniMax M3

Open Weight

License

Celeris-1

Proprietary

MiniMax M3

Open Weight

Release date

Celeris-1

2026-07-22

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
The public evidence has no benchmark result shared by both models, so it does not support a quality verdict.
Workload cost
Repository review: $0.0121 vs $0.0186. Cache-heavy agent loop: $0.051 vs $0.03.
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 evidence31 rows

Agentic

  • Terminal-Bench 2.0

    Celeris-1—
    MiniMax M366%
    Source

    Not directly comparable

  • BrowseComp

    Celeris-1—
    MiniMax M383.5%
    Source

    Not directly comparable

  • OSWorld-Verified

    Celeris-1—
    MiniMax M370.1%
    Source

    Not directly comparable

  • MCP Atlas

    Celeris-1—
    MiniMax M374.2%
    Source

    Not directly comparable

  • Claw-Eval

    Celeris-1—
    MiniMax M374.5%
    Source

    Not directly comparable

  • BankerToolBench

    Celeris-1—
    MiniMax M376.1%
    Source

    Not directly comparable

  • ResearchClawBench

    Celeris-1—
    MiniMax M319.8%
    Source

    Not directly comparable

  • OSWorld 2.0

    Celeris-1—
    MiniMax M34.6%
    Source

    Not directly comparable

  • Terminal-Bench 2.1 (Vals)

    Celeris-1—
    MiniMax M353.6%
    Source

    Not directly comparable

Coding

  • SWE-bench Verified

    Celeris-1—
    MiniMax M380.5%
    Source

    Not directly comparable

  • SWE-bench Pro

    Celeris-1—
    MiniMax M359%
    Source

    Not directly comparable

  • Terminal-Bench 2.0

    Celeris-1—
    MiniMax M366.0%
    Source

    Not directly comparable

  • NL2Repo

    Celeris-1—
    MiniMax M342.1%
    Source

    Not directly comparable

  • VIBE V2

    Celeris-1—
    MiniMax M350.1%
    Source

    Not directly comparable

  • SVG-Bench

    Celeris-1—
    MiniMax M363.7%
    Source

    Not directly comparable

  • KernelBench Hard

    Celeris-1—
    MiniMax M328.8%
    Source

    Not directly comparable

  • OpenHarmony Bench

    Celeris-1—
    MiniMax M348.4%
    Source

    Not directly comparable

  • LiveCodeBench (Vals)

    Celeris-1—
    MiniMax M382.2%
    Source

    Not directly comparable

  • SWE-bench (Vals)

    Celeris-1—
    MiniMax M375.0%
    Source

    Not directly comparable

Reasoning

  • DROP

    Celeris-181.4%
    Source
    MiniMax M3—

    Not directly comparable

Knowledge

  • MMLU-Pro

    Celeris-175.9%
    Source
    MiniMax M3—

    Not directly comparable

  • GPQA Diamond (Vals)

    Celeris-1—
    MiniMax M392.7%
    Source

    Not directly comparable

  • MMLU-Pro (Vals)

    Celeris-1—
    MiniMax M384.2%
    Source

    Not directly comparable

Math

  • GSM8K

    Celeris-193.7%
    Source
    MiniMax M3—

    Not directly comparable

  • USAMO 2026

    Celeris-1—
    MiniMax M385.7%
    Source

    Not directly comparable

Multimodal

  • OfficeQA Pro

    Celeris-1—
    MiniMax M345.1%
    Source

    Not directly comparable

  • OmniDocBench 1.5

    Celeris-1—
    MiniMax M391.6%
    Source

    Not directly comparable

  • MMMU-Pro

    Celeris-1—
    MiniMax M378.1%
    Source

    Not directly comparable

  • VideoMMMU

    Celeris-1—
    MiniMax M384.6%
    Source

    Not directly comparable

  • Video-MME (with subtitle)

    Celeris-1—
    MiniMax M385.4%
    Source

    Not directly comparable

Instruction following

  • IFEval

    Celeris-180.8%
    Source
    MiniMax M3—

    Not directly comparable

Frequently asked questions

Which is better, Celeris-1 or MiniMax M3?

The public evidence has no benchmark result shared by both models, so it does not support a quality verdict. The page therefore keeps the decision tied to the specific documented workload.

Which is better for coding, Celeris-1 or MiniMax M3?

Celeris-1 is not ranked on the public lane for coding, so no winner is named for coding.

Which is better for agentic tasks, Celeris-1 or MiniMax M3?

Celeris-1 is not ranked on the public lane for agentic tasks, so no winner is named for agentic tasks.

Which costs less, Celeris-1 or MiniMax M3?

For the stated presets, chat costs $0.00055 on Celeris-1 and $0.0009 on MiniMax M3; repository review costs $0.0121 and $0.0186; the cache-heavy agent loop costs $0.051 and $0.03. Celeris-1 does not fit this workload in one request. Celeris-1 has no published cached-input rate, so cached tokens use its listed input rate.

Which has the larger context window, Celeris-1 or MiniMax M3?

MiniMax M3 has the larger documented context window: 1M, compared with 131,072 tokens.

Related comparisons

Compare API pricing Read the methodology Open the model selector

Last updated September 4, 2026

Watch Celeris-1 vs MiniMax M3

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

Read a sample issue

Join 2,000+ readers.