Enter opens the highlighted result. Esc closes.

Qwen3.8 27B vs MiMo-V2.6-Flash

Updated October 21, 2026. 57 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace Qwen3.8 27B Replace MiMo-V2.6-Flash
Qwen3.8 27BQwen
Coding
73.6
Overall
33.7
Output $/M
$1.35
12
MiMo-V2.6-FlashXiaomi
Coding
61.9
Overall
37.9
Output $/M
$0.28
45
vs boards won of 57 shared

Decision reading

Pick Qwen3.8 27B for coding: it scores 73.6 on the Vibe Coding Index against 61.9 for MiMo-V2.6-Flash.

Pick MiMo-V2.6-Flash when output tokens cost 4.8x less, or when the context window is larger, 1.05M against 1M tokens.

Model AQwen3.8 27BQwen73.6/1002 independent benchmarks$1.26 per solved taskModel BMiMo-V2.6-FlashXiaomi61.9/1001 independent benchmarkNo cost per solved task
Shared results
57
Qwen3.8 27B only
28
MiMo-V2.6-Flash only
9
Like-for-like categories
7 of 9

Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
Qwen3.8 27B 73.6 from 2 benchmarks, MiMo-V2.6-Flash 61.9 from 1 benchmark.
Widest category
Math, 18.0 points apart on average over 3 shared benchmarks.
Widest single gap
ProofBench: Qwen3.8 27B 16.0% against MiMo-V2.6-Flash 63.0%, 47.0 points apart on a 0 to 100 scale.
Output price
Qwen3.8 27B $1.35 against MiMo-V2.6-Flash $0.28 per million output tokens, 4.8x apart.
Context window
Qwen3.8 27B takes 1M tokens, MiMo-V2.6-Flash 1.05M.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryQwen3.8 27BBarsMiMo-V2.6-FlashTies
Coding013
Intelligence01
Agentic work14
Reasoning24
Math03
Knowledge23
General717
All shared boards1245
CodingLike-for-likeFour of these boards feed the Vibe Coding Index013
Coding benchmarks for Qwen3.8 27B and MiMo-V2.6-Flash
BenchmarkQwen3.8 27BMiMo-V2.6-FlashLeader
LiveBench coding86.8%Not measuredOnly Qwen3.8 27B measured
LMArena WebDev15931634MiMo-V2.6-Flash
AA Coding Index68.1Not measuredOnly Qwen3.8 27B measured
LiveCodeBench (Vals AI)84.0%Not measuredOnly Qwen3.8 27B measured
SWE-bench Verified (Vals AI)86.0%Not measuredOnly Qwen3.8 27B measured
Terminal-Bench 4 (Vals AI)Not measured24.2%Only MiMo-V2.6-Flash measured
Vibe Code Bench (Vals AI)64.8%79.0%MiMo-V2.6-Flash
Code migration (Vals AI)14.2%40.9%MiMo-V2.6-Flash
ProgramBench (Vals AI)0.0%0.5%MiMo-V2.6-Flash
IOI (Vals AI)39.1%47.7%MiMo-V2.6-Flash
Terminal-Bench 2.1 (Vals AI)58.4%76.4%MiMo-V2.6-Flash
SkillsBench (Vals AI)38.1%Not measuredOnly Qwen3.8 27B measured
CyberBench (Vals AI)67.8%75.4%MiMo-V2.6-Flash
SciCode46.6%51.3%MiMo-V2.6-Flash
ALE-BenchNot measured1244.2Only MiMo-V2.6-Flash measured
LMArena Text: coding14831499MiMo-V2.6-Flash
LMArena Text: software and IT14781497MiMo-V2.6-Flash
LMArena WebDev: React16021648MiMo-V2.6-Flash
LMArena WebDev: HTML15701641MiMo-V2.6-Flash
LMArena WebDev: from an image15771596MiMo-V2.6-Flash
Neither model has a result on the other 18 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
Intelligence benchmarks for Qwen3.8 27B and MiMo-V2.6-Flash
BenchmarkQwen3.8 27BMiMo-V2.6-FlashLeader
AA Intelligence Index33.737.9MiMo-V2.6-Flash
Epoch Capabilities Index151.7Not measuredOnly Qwen3.8 27B measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index14
Agentic work benchmarks for Qwen3.8 27B and MiMo-V2.6-Flash
BenchmarkQwen3.8 27BMiMo-V2.6-FlashLeader
AA Agentic Index45.8Not measuredOnly Qwen3.8 27B measured
LiveBench research59.4%Not measuredOnly Qwen3.8 27B measured
LiveBench browser32.0%Not measuredOnly Qwen3.8 27B measured
LiveBench long agentic27.9%Not measuredOnly Qwen3.8 27B measured
SREBench (Vals AI)Not measured4.2%Only MiMo-V2.6-Flash measured
Finance Agent (Vals AI)48.5%56.3%MiMo-V2.6-Flash
Tax Agent Bench (Vals AI)30.9%28.5%Qwen3.8 27B
Legal Research Bench (Vals AI)36.1%38.0%MiMo-V2.6-Flash
Harvey's Legal Agent Benchmark (Vals AI)Not measured13.3%Only MiMo-V2.6-Flash measured
Excel Modeling Benchmark (Vals AI)59.7%65.5%MiMo-V2.6-Flash
BioMysteryBench (Vals AI)Not measured69.3%Only MiMo-V2.6-Flash measured
Terminal-Bench-Science (Vals AI)0.0%4.3%MiMo-V2.6-Flash
Vals Science IndexNot measured31.7%Only MiMo-V2.6-Flash measured
APEX-Agents47.5%Not measuredOnly Qwen3.8 27B measured
Neither model has a result on the other 25 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index24
Reasoning benchmarks for Qwen3.8 27B and MiMo-V2.6-Flash
BenchmarkQwen3.8 27BMiMo-V2.6-FlashLeader
GPQA Diamond88.9%Not measuredOnly Qwen3.8 27B measured
LiveBench reasoning80.0%Not measuredOnly Qwen3.8 27B measured
MysteryMechanism (Vals AI)Not measured21.6%Only MiMo-V2.6-Flash measured
MMMU Pro (Vals AI)83.9%Not measuredOnly Qwen3.8 27B measured
GPQA Diamond (Artificial Analysis run)90.5%Not measuredOnly Qwen3.8 27B measured
Humanity's Last Exam (Artificial Analysis run)33.9%35.1%MiMo-V2.6-Flash
ARC-AGI87.5%Not measuredOnly Qwen3.8 27B measured
ARC-AGI-242.4%Not measuredOnly Qwen3.8 27B measured
LMArena Text: hard prompts14591475MiMo-V2.6-Flash
LMArena Text: expert14791496MiMo-V2.6-Flash
LMArena Text: hard prompts in English14611493MiMo-V2.6-Flash
LMArena Vision: diagrams12851276Qwen3.8 27B
LMArena Vision: homework12951283Qwen3.8 27B
Neither model has a result on the other 8 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index03
Math benchmarks for Qwen3.8 27B and MiMo-V2.6-Flash
BenchmarkQwen3.8 27BMiMo-V2.6-FlashLeader
ProofBench16.0%63.0%MiMo-V2.6-Flash
LMArena Text: math14551476MiMo-V2.6-Flash
LMArena Text: mathematical work14581478MiMo-V2.6-Flash
Neither model has a result on the other 10 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index23
Knowledge benchmarks for Qwen3.8 27B and MiMo-V2.6-Flash
BenchmarkQwen3.8 27BMiMo-V2.6-FlashLeader
MMLU Pro84.3%Not measuredOnly Qwen3.8 27B measured
LegalBench (Vals AI)82.4%Not measuredOnly Qwen3.8 27B measured
TaxEval (Vals AI)70.8%Not measuredOnly Qwen3.8 27B measured
MortgageTax (Vals AI)64.9%Not measuredOnly Qwen3.8 27B measured
MedCode (Vals AI)28.7%41.1%MiMo-V2.6-Flash
LMArena Text: science14671462Qwen3.8 27B
LMArena Text: medicine and healthcare14371439MiMo-V2.6-Flash
LMArena Text: legal and government14471441Qwen3.8 27B
LMArena Text: business and finance14381454MiMo-V2.6-Flash
Neither model has a result on the other 8 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index1 board, none shared
Web design benchmarks for Qwen3.8 27B and MiMo-V2.6-Flash
BenchmarkQwen3.8 27BMiMo-V2.6-FlashLeader
LiveBench design75.9%Not measuredOnly Qwen3.8 27B measured
Neither model has a result on the other 8 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index717
General benchmarks for Qwen3.8 27B and MiMo-V2.6-Flash
BenchmarkQwen3.8 27BMiMo-V2.6-FlashLeader
Vals IndexNot measured54.8%Only MiMo-V2.6-Flash measured
LiveBench data analysis76.6%Not measuredOnly Qwen3.8 27B measured
Public Benefits Bench (Vals AI)Not measured67.6%Only MiMo-V2.6-Flash measured
SAGE (Vals AI)52.4%43.5%Qwen3.8 27B
MedScribe (Vals AI)83.8%85.3%MiMo-V2.6-Flash
Long context reasoning (AA-LCR)82.0%74.3%Qwen3.8 27B
DTBench88.0%Not measuredOnly Qwen3.8 27B measured
LMCA41.4%Not measuredOnly Qwen3.8 27B measured
Surface Evolver Bench45.0%Not measuredOnly Qwen3.8 27B measured
LMArena Text14401453MiMo-V2.6-Flash
LMArena Text: instruction following14361452MiMo-V2.6-Flash
LMArena Text: creative writing13831403MiMo-V2.6-Flash
LMArena Vision12691256Qwen3.8 27B
LMArena Text: English14481476MiMo-V2.6-Flash
LMArena Text: non-English14281431MiMo-V2.6-Flash
LMArena Text: Chinese15011509MiMo-V2.6-Flash
LMArena Text: German14431458MiMo-V2.6-Flash
LMArena Text: Japanese13871359Qwen3.8 27B
LMArena Text: Korean13841425MiMo-V2.6-Flash
LMArena Text: French14651480MiMo-V2.6-Flash
LMArena Text: Spanish14491453MiMo-V2.6-Flash
LMArena Text: Russian14151409Qwen3.8 27B
LMArena Text: Polish14391455MiMo-V2.6-Flash
LMArena Text: multi-turn14401447MiMo-V2.6-Flash
LMArena Text: longer queries14481459MiMo-V2.6-Flash
LMArena Text: writing and language14051424MiMo-V2.6-Flash
LMArena Text: entertainment and media13931415MiMo-V2.6-Flash
LMArena Vision: English12671269MiMo-V2.6-Flash
LMArena Vision: Chinese1326Not measuredOnly Qwen3.8 27B measured
LMArena Vision: reading text12811270Qwen3.8 27B
LMArena Vision: humor1235Not measuredOnly Qwen3.8 27B measured
LMArena Vision: creative writing12371231Qwen3.8 27B
Neither model has a result on the other 14 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Qwen3.8 27B and MiMo-V2.6-Flash
BenchmarkQwen3.8 27BMiMo-V2.6-FlashLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

Qwen3.8 27B$0.0007Fits in one request
MiMo-V2.6-Flash$0.0003Fits in one request

MiMo-V2.6-Flash costs less

Repository review

50K input + 3K output tokens

Qwen3.8 27B$0.0060Fits in one request
MiMo-V2.6-Flash$0.0078Fits in one request

Qwen3.8 27B costs less

Agent loop

200K input + 10K output tokens

Qwen3.8 27B$0.021Fits in one request
MiMo-V2.6-Flash$0.031Fits in one request

Qwen3.8 27B costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of Qwen3.8 27B and MiMo-V2.6-Flash
SpecificationQwen3.8 27BXiaomi MiMo-V2.6-Flash
MakerQwenXiaomi
WeightsOpen weightsOpen weights
ReleasedAugust 14, 2026September 21, 2026
Context window1M1.05M
Input $/Mtok$0.04$0.14
Output $/Mtok$1.35$0.28
Coding index73.661.9
Independent measurements behind it21
Cost per solved task$1.26Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Qwen3.8 27B better than MiMo-V2.6-Flash for coding?

Pick Qwen3.8 27B for coding: it scores 73.6 on the Vibe Coding Index against 61.9 for MiMo-V2.6-Flash. Pick MiMo-V2.6-Flash when output tokens cost 4.8x less, or when the context window is larger, 1.05M against 1M tokens.

Which is cheaper, Qwen3.8 27B or MiMo-V2.6-Flash?

Qwen3.8 27B costs $0.04 per million input tokens and $1.35 per million output tokens; MiMo-V2.6-Flash costs $0.14 and $0.28. Cost per solved task is known for Qwen3.8 27B only.

Which has the bigger context window, Qwen3.8 27B or MiMo-V2.6-Flash?

MiMo-V2.6-Flash, with 1.05M tokens against 1M.

How many benchmarks is the Qwen3.8 27B vs MiMo-V2.6-Flash comparison based on?

57 benchmarks have a result for both models as of October 21, 2026. Qwen3.8 27B's coding index rests on 2 independent benchmarks and MiMo-V2.6-Flash's on 1. Where only one model has a result, the page says so instead of naming a winner.

Compare next