Why is this comparison indexable?

It has 6 verifiable comparison points, and both models have pricing or benchmark data.

Are missing metrics invented?

No. Metrics without LMSpeed data are omitted from this report.

GitHub

Model compare	DeepSeek V4 Flash	MiMo-V2.5-Pro
Overall leader	Contender	Leading
Weighted overall score	16.0 pts	84.0 pts
Benchmark category leads	0 categories	5 categories
Operational advantages	Cheapest input price, Average speed, Free providers, Provider coverage	First-token latency
Context window	1.0M tokens	1.1M tokens
Max output	393.2K tokens	131.1K tokens
Modalities	Input Text Output Text	Input Text

Model compare	DeepSeek V4 Flash	MiMo-V2.5-Pro
Developer	DeepSeek	Xiaomi
Released	Apr 2026	Apr 2026
Parameters	No data	No data
Tokenizer	DeepSeek	Other
Knowledge cutoff	No data	No data
OpenRouter ID	deepseek/deepseek-v4-flash	xiaomi/mimo-v2.5-pro
References	No data	No data

Aggregatereported

BenchLM overall score

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

56.0

Rank #41/84 · confidence 2 · eval date 2026-04-24

+3.0

MiMo-V2.5-Pro

winner

59.0

Rank #31/84 · confidence 1 · eval date 2026-04-22

Pricingverified

Output price

SourceDeepSeek V4 Flash (Reasoning, High Effort)

DeepSeek V4 Flash

winner

$0.280/M

Rank #6/162 · confidence 4

+$0.590/M

MiMo-V2.5-Pro

$0.870/M

Rank #34/162 · confidence 4

Pricingverified

Blended price

SourceDeepSeek V4 Flash (Reasoning, High Effort)

DeepSeek V4 Flash

winner

$0.175/M

Rank #11/162 · confidence 4

+$0.369/M

MiMo-V2.5-Pro

$0.544/M

Rank #47/162 · confidence 4

Pricingverified

Input price

SourceDeepSeek V4 Flash (Reasoning, High Effort)

DeepSeek V4 Flash

winner

$0.140/M

Rank #17/162 · confidence 4

+$0.295/M

MiMo-V2.5-Pro

$0.435/M

Rank #66/162 · confidence 4

Reasoningreported

AA-LCR

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

63.0

Rank #53/89 · confidence 2 · eval date 2026-04-24

+10.3

MiMo-V2.5-Pro

winner

73.3

Rank #12/89 · confidence 1 · eval date 2026-04-22

Reasoningverified

HLE

SourceDeepSeek V4 Flash (Reasoning, High Effort)

DeepSeek V4 Flash

27.8%

Rank #36/187 · confidence 4

+6.0%

MiMo-V2.5-Pro

winner

33.8%

Rank #23/187 · confidence 4

Reasoningreported

CritPt

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

winner

7.1

Rank #29/73 · confidence 2 · eval date 2026-04-24

+3.1

MiMo-V2.5-Pro

4.0

Rank #38/73 · confidence 1 · eval date 2026-04-22

Reasoningverified

GPQA

SourceDeepSeek V4 Flash (Reasoning, High Effort)

DeepSeek V4 Flash

winner

86.7%

Rank #36/188 · confidence 4

+0.1%

MiMo-V2.5-Pro

86.6%

Rank #39/188 · confidence 4

Codingreported

Terminal-Bench Hard

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

35.6

Rank #12/14 · confidence 2 · eval date 2026-04-24

+7.6

MiMo-V2.5-Pro

winner

43.2

Rank #10/14 · confidence 1 · eval date 2026-04-22

Codingreported

Terminal-Bench 2.0

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

56.9

Rank #22/26 · confidence 2 · eval date 2026-04-24

+11.5

MiMo-V2.5-Pro

winner

68.4

Rank #14/26 · confidence 1 · eval date 2026-04-22

Codingverified

SciCode

SourceDeepSeek V4 Flash (Reasoning, High Effort)

DeepSeek V4 Flash

42.0%

Rank #52/185 · confidence 4

+8.2%

MiMo-V2.5-Pro

winner

50.2%

Rank #16/185 · confidence 4

Codingreported

AA Coding Index

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

56.2

Rank #23/59 · confidence 2 · eval date 2026-04-24

+4.0

MiMo-V2.5-Pro

winner

60.2

Rank #18/59 · confidence 1 · eval date 2026-04-22

Codingreported

SWE-bench Pro

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

52.6

Rank #31/37 · confidence 2 · eval date 2026-04-24

+4.6

MiMo-V2.5-Pro

winner

57.2

Rank #18/37 · confidence 1 · eval date 2026-04-22

Codingreported

AA-SciCode

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

44.9

Rank #41/96 · confidence 2 · eval date 2026-04-24

+5.3

MiMo-V2.5-Pro

winner

50.2

Rank #20/96 · confidence 1 · eval date 2026-04-22

Codingreported

BenchLM Coding score

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

winner

46.8

Rank #47/65 · confidence 2 · eval date 2026-04-24

+7.4

MiMo-V2.5-Pro

39.4

Rank #56/65 · confidence 1 · eval date 2026-04-22

Knowledgereported

AA Openness Index

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

winner

50.0

Rank #1/8 · confidence 2 · eval date 2026-04-24

+11.1

MiMo-V2.5-Pro

38.9

Rank #4/8 · confidence 1 · eval date 2026-04-22

Knowledgereported

AA-Omniscience Hallucination Rate

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

winner

95.8

Rank #1/90 · confidence 2 · eval date 2026-04-24

+71.3

MiMo-V2.5-Pro

24.5

Rank #87/90 · confidence 1 · eval date 2026-04-22

Knowledgereported

Artificial Analysis Intelligence Index

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

40.3

Rank #27/99 · confidence 2 · eval date 2026-04-24

+2.0

MiMo-V2.5-Pro

winner

42.2

Rank #21/99 · confidence 1 · eval date 2026-04-22

Knowledgereported

AA-HLE

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

32.1

Rank #27/96 · confidence 2 · eval date 2026-04-24

+1.7

MiMo-V2.5-Pro

winner

33.8

Rank #23/96 · confidence 1 · eval date 2026-04-22

Knowledgereported

BenchLM Knowledge score

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

60.1

Rank #40/55 · confidence 2 · eval date 2026-04-24

+10.6

MiMo-V2.5-Pro

winner

70.7

Rank #25/55 · confidence 1 · eval date 2026-04-22

Knowledgereported

AA-GPQA Diamond

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

winner

89.4

Rank #27/96 · confidence 2 · eval date 2026-04-24

+2.8

MiMo-V2.5-Pro

86.6

Rank #39/96 · confidence 1 · eval date 2026-04-22

Knowledgereported

AA-Omniscience Accuracy

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

winner

37.2

Rank #33/90 · confidence 2 · eval date 2026-04-24

+14.6

MiMo-V2.5-Pro

22.6

Rank #67/90 · confidence 1 · eval date 2026-04-22

Multimodalreported

Design Arena Website

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

1238.0

Rank #33/66 · confidence 2 · eval date 2026-04-24

+60.0

MiMo-V2.5-Pro

winner

1298.0

Rank #14/66 · confidence 1 · eval date 2026-04-22

Instruction followingreported

AA-IFBench

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

79.2

Rank #5/86 · confidence 2 · eval date 2026-04-24

+0.7

MiMo-V2.5-Pro

winner

79.9

Rank #4/86 · confidence 1 · eval date 2026-04-22

Agentsreported

Claw-Eval

SourceDeepSeek V4 Flash

DeepSeek V4 Flash

57.8

Rank #17/28 · confidence 2 · eval date 2026-04-24

+6.0

MiMo-V2.5-Pro

winner

63.8

Rank #8/28 · confidence 1 · eval date 2026-04-22

Agentsreported

Gert Labs

SourceDeepSeek V4 Flash

DeepSeek V4 Flash

54.4

Rank #18/50 · confidence 2 · eval date 2026-04-24

+8.4

MiMo-V2.5-Pro

winner

62.7

Rank #9/50 · confidence 1 · eval date 2026-04-22

Agentsreported

AA ITBench

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

31.5

Rank #11/13 · confidence 2 · eval date 2026-04-24

+6.7

MiMo-V2.5-Pro

winner

38.2

Rank #10/13 · confidence 1 · eval date 2026-04-22

Agentsreported

AA Briefcase

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

833.0

Rank #13/13 · confidence 2 · eval date 2026-04-24

+45.0

MiMo-V2.5-Pro

winner

878.0

Rank #12/13 · confidence 1 · eval date 2026-04-22

Agentsreported

AA Harvey LAB

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

winner

81.3

Rank #14/17 · confidence 2 · eval date 2026-04-24

+8.0

MiMo-V2.5-Pro

73.3

Rank #15/17 · confidence 1 · eval date 2026-04-22

Agentsreported

AA Agentic Index

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

winner

31.1

Rank #16/55 · confidence 2 · eval date 2026-04-24

+1.9

MiMo-V2.5-Pro

29.1

Rank #24/55 · confidence 1 · eval date 2026-04-22

Agentsreported

GDPval-AA

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

1189.0

Rank #21/55 · confidence 2 · eval date 2026-04-24

+76.0

MiMo-V2.5-Pro

winner

1265.0

Rank #17/55 · confidence 1 · eval date 2026-04-22

Agentsreported

GDPval-AA

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

34.4

Rank #21/50 · confidence 2 · eval date 2026-04-24

+3.9

MiMo-V2.5-Pro

winner

38.3

Rank #17/50 · confidence 1 · eval date 2026-04-22

Agentsreported

Terminal-Bench 2.0

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

56.9

Rank #32/46 · confidence 2 · eval date 2026-04-24

+11.5

MiMo-V2.5-Pro

winner

68.4

Rank #17/46 · confidence 1 · eval date 2026-04-22

Agentsreported

τ²-bench results

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

winner

95.0

Rank #19/84 · confidence 2 · eval date 2026-04-24

+0.8

MiMo-V2.5-Pro

94.2

Rank #23/84 · confidence 1 · eval date 2026-04-22

Agentsreported

BenchLM Agentic score

SourceDeepSeek V4 Flash (Max)

DeepSeek V4 Flash

46.7

Rank #37/53 · confidence 2 · eval date 2026-04-24

+15.5

MiMo-V2.5-Pro

winner

62.2

Rank #26/53 · confidence 1 · eval date 2026-04-22

Provider	DeepSeek V4 Flash	MiMo-V2.5-Pro
70 tests	DeepSeek V4 Flash speed / latency 59 tok/s / 29380ms input / output No data	MiMo-V2.5-Pro speed / latency N/A / N/A input / output No data
25 tests	DeepSeek V4 Flash speed / latency

Comparison sheet

Model metadata

When to choose each model

Benchmark score comparison

Category performance

Agents

Coding

Reasoning

Knowledge

Math

Multilingual

Multimodal

Instruction following

Professional benchmark details

BenchLM overall score

Output price

Blended price

Input price

AA-LCR

HLE

CritPt

GPQA

Terminal-Bench Hard

Terminal-Bench 2.0

SciCode

AA Coding Index

SWE-bench Pro

AA-SciCode

BenchLM Coding score

AA Openness Index

AA-Omniscience Hallucination Rate

Artificial Analysis Intelligence Index

AA-HLE

BenchLM Knowledge score

AA-GPQA Diamond

AA-Omniscience Accuracy

Design Arena Website

AA-IFBench

Claw-Eval

Gert Labs

AA ITBench

AA Briefcase

AA Harvey LAB

AA Agentic Index

GDPval-AA

GDPval-AA

Terminal-Bench 2.0

τ²-bench results

BenchLM Agentic score

API audit comparison

Provider examples

FAQ

Related compare reports