Skip to main content

Qwen 3.8 27Bvs.Mistral Large 4

Which one is better?

Mistral Large 4 is clearly ahead in the overall index, but costs 34 times (input) and 1.7 times (output) as much per token.

IntelligenceOverall index, 0 to 100 · higher is better
11

Qwen 3.8 27B

Rank 44 to 57

28

Mistral Large 4

Rank 34 to 43

SpeedTokens per second · higher is better
40

Qwen 3.8 27B

slow

113

Mistral Large 4

fairly fast

PriceDollars per 1M tokens · lower is better
$0.66

Qwen 3.8 27B

fairly cheap

$2.07

Mistral Large 4

fairly expensive

Qwen 3.8 27B

Subscription

No Alibaba subscription includes this model. Through other providers from ≈ €1.06 a month (Command Code Go, with limited credits). 10 plans in Germany.

Via API

$0.04 input, $2.50 output per 1M tokens according to the provider; the price index has listed it at $0.42 and $3 since Oct 27, 2026.

Mistral Large 4

Subscription

No Mistral AI subscription includes this model. Through other providers from ≈ €1.06 a month (Command Code Go, with limited credits). 12 plans in Germany.

Via API

$1.36 input, $4.18 output per 1M tokens.

Overall index with data through Oct 29, 2026 · Prices from Oct 29, 2026 · Speed: measured by Artificial Analysis · Subscriptions as of Oct 24, 2026 · Subscription prices incl. VAT

Where is the difference?

Values from 0 to 100 per task area. We only name a winner when the gap is large enough and there are enough tests.

Qwen 3.8 27B ahead: 0Mistral Large 4 ahead: 7level: 1too little data: 2of 10 areas

Qwen 3.8 27BMistral Large 40 to 100, higher is better
Areas in the overall index
CodingMistral Large 4 aheadQwen 3.8 27B: 13Mistral Large 4: 38
ReasoningMistral Large 4 aheadQwen 3.8 27B: 6Mistral Large 4: 26
Mathematicstoo little dataQwen 3.8 27B: 8Mistral Large 4: 10
KnowledgeMistral Large 4 aheadQwen 3.8 27B: 5Mistral Large 4: 11
AgentsMistral Large 4 aheadQwen 3.8 27B: 21Mistral Large 4: 48
Generalonly Mistral Large 4 has a scoreQwen 3.8 27B: not enough dataMistral Large 4: 28
MultimodalMistral Large 4 aheadQwen 3.8 27B: 13Mistral Large 4: 23
Specialist fields
FinanceMistral Large 4 aheadQwen 3.8 27B: 34Mistral Large 4: 54
LegalMistral Large 4 aheadQwen 3.8 27B: 23Mistral Large 4: 49
HealthlevelQwen 3.8 27B: 3Mistral Large 4: 18
Show 1 more areas that do not count toward the overall index
Qwen 3.8 27BMistral Large 40 to 100, higher is better
Cybersecurityonly Qwen 3.8 27B has a scoreQwen 3.8 27B: 19Mistral Large 4: not enough data

See for yourself: same task, one attempt

SVG short test

One SVG task with the same fixed settings for every model, without hidden thinking.

Shared test round: 43b8c5c46830

Qwen 3.8 27B

Qwen 3.8 27B: SVG short test
Test run: Oct 14, 2026 · alibaba
Test detailsAPI reports 0 reasoning tokensAPI model: qwen/qwen3.8-27b865 output tokens

Mistral Large 4

Mistral Large 4: SVG short test
Test run: Oct 14, 2026 · mistral
Test detailsAPI reports 0 reasoning tokensAPI model: mistralai/mistral-large-4-0564 output tokens
Earlier tests

This is the lowest documented setting for each provider. It does not make compute comparable across models.

Pelican on a bicycle

A community classic for free-form SVG drawing.

Qwen 3.8 27B

The response reached the token limit. No complete recording is available.

Reasoning (requested): disabled (enabled: false)

Current attempt
Run date
Oct 11, 2026
API provider
Alibaba
Requested API model
qwen/qwen3.8-27b
API model reported in response
qwen/qwen3.8-27b
Reasoning setting
disabled (enabled: false)
Pinned endpoint
Alibaba | qwen/qwen3.8-27b-20260814
Temperature
Provider default
Token limit including reasoning
8,192 tokens
Reasoning tokens
0
Test protocol
community-visual-low-v1
Technical checks
Failed
Response time
119.1 s

Not yet visually reviewed. A successful recording does not confirm correct physics or full compliance with the prompt.

Mistral Large 4

Mistral Large 4: Pelican on a bicycle
A community classic for free-form SVG drawing.

Reasoning (requested): none

Current attempt
Run date
Oct 11, 2026
API provider
Mistral
Requested API model
mistralai/mistral-large-4-0
API model reported in response
mistralai/mistral-large-4-0
Reasoning setting
none
Pinned endpoint
Mistral | mistralai/mistral-large-4-0-20261006
Temperature
Provider default
Token limit including reasoning
8,192 tokens
Reasoning tokens
0
Test protocol
community-visual-low-v1
Technical checks
Passed, not a quality rating
Response time
28.7 s

Visually reviewed. This individual test does not establish overall model quality.

Prompt and test conditions

Generate an SVG of a pelican riding a bicycle

Token limit including reasoning: 8,192 tokens

Task origin (Simon Willison)

All data side by side

The better value is bold. For the overall index only when the gap is clear.

MetricQwen 3.8 27BMistral Large 4
Intelligence
Overall index0 to 100, higher is better1128
RankRank 44 to 57 of 95Rank 34 to 43 of 95
Price per 1M tokens
Inputwhat you send$0.04$1.36
Outputwhat the model writes$2.50$4.18
Speed
Tokens per second40113
Technical
Context windowhow much text the model reads at once262,144 tokens≈ 390 A4 pages524,288 tokens≈ 790 A4 pages
Knowledge cutoffn/an/a
ReleasedAug 14, 2026Oct 6, 2026
AvailabilityActivePreview
Open weightsyesn/a
  • Mistral Large 4, Launch pricing announced October 6, 2026: Mistral announced a 50% discount for two weeks. The published introductory rates are $0.68 input, $2.09 output and $0.07 cache read per 1M tokens. Price comparisons use the standard rates. Source

Prices without caching or batch discounts · Page count as a rule of thumb from Artificial Analysis

Shared benchmarks

Single results from the same benchmark version, task and metric. A value is bold only with the same test setup and a clear gap.

A: Qwen 3.8 27BB: Mistral Large 4

Code MigrationOverall · higher is betterTest setup not documented as identicalQwen 3.8 27B: 14.2%Mistral Large 4: 30.6%Vals AIAs of 10/18/2026
EMBOverall · higher is betterTest setup not documented as identicalQwen 3.8 27B: 59.7%Mistral Large 4: 56.2%Vals AIAs of 10/20/2026
Finance Agent (v2)Overall · higher is betterTest setup not documented as identicalQwen 3.8 27B: 48.6%Mistral Large 4: 54.7%Vals AIAs of 10/20/2026
IOIOverall · higher is betterTest setup not documented as identicalQwen 3.8 27B: 39.1%Mistral Large 4: 45.3%Vals AIAs of 10/24/2026
Legal Research BenchOverall · All-pass · higher is betterTest setup not documented as identicalQwen 3.8 27B: 36.1%Mistral Large 4: 31.7%Vals AIAs of 10/26/2026
MedCodeOverall · higher is betterTest setup not documented as identicalQwen 3.8 27B: 28.7%Mistral Large 4: 40.7%Vals AIAs of 10/13/2026
MedScribeOverall · higher is betterTest setup not documented as identicalQwen 3.8 27B: 83.8%Mistral Large 4: 80.4%Vals AIAs of 10/13/2026
ProofBench v1.1Overall · higher is betterTest setup not documented as identicalQwen 3.8 27B: 16%Mistral Large 4: 10%Vals AIAs of 10/24/2026
Tax Agent BenchOverall · higher is betterTest setup not documented as identicalQwen 3.8 27B: 30.9%Mistral Large 4: 30.4%Vals AIAs of 10/26/2026
Vibe Code Bench v1.1Overall · higher is betterTest setup not documented as identicalQwen 3.8 27B: 64.9%Mistral Large 4: 78.4%Vals AIAs of 10/26/2026
Show 54 more benchmarks
Code MigrationCLI · higher is betterTest setup not documented as identicalQwen 3.8 27B: 2.2%Mistral Large 4: 17.4%Vals AIAs of 10/18/2026
EMBComps · higher is betterTest setup not documented as identicalQwen 3.8 27B: 53.3%Mistral Large 4: 56.6%Vals AIAs of 10/20/2026
Finance Agent (v2)Adjustments · higher is betterTest setup not documented as identicalQwen 3.8 27B: 47.3%Mistral Large 4: 52.4%Vals AIAs of 10/20/2026
IOIIOI 2024 · higher is betterTest setup not documented as identicalQwen 3.8 27B: 31.8%Mistral Large 4: 50.2%Vals AIAs of 10/24/2026
Legal Research BenchAdministrative / Regulatory · All-pass · higher is betterTest setup not documented as identicalQwen 3.8 27B: 45.5%Mistral Large 4: 34.8%Vals AIAs of 10/26/2026
Tax Agent BenchControversy & Precedence Analysis · higher is betterTest setup not documented as identicalQwen 3.8 27B: 6.7%Mistral Large 4: 3.3%Vals AIAs of 10/26/2026
Code MigrationCOBOL · higher is betterTest setup not documented as identicalQwen 3.8 27B: 50%Mistral Large 4: 70%Vals AIAs of 10/18/2026
EMBDataroom · higher is betterTest setup not documented as identicalQwen 3.8 27B: 79.7%Mistral Large 4: 75%Vals AIAs of 10/20/2026
Finance Agent (v2)All-Pass · higher is betterTest setup not documented as identicalQwen 3.8 27B: 36.8%Mistral Large 4: 43.8%Vals AIAs of 10/20/2026
IOIIOI 2025 · higher is betterTest setup not documented as identicalQwen 3.8 27B: 48.5%Mistral Large 4: 37.8%Vals AIAs of 10/24/2026
Legal Research BenchAdministrative / Regulatory · Weighted partial credit · higher is betterTest setup not documented as identicalQwen 3.8 27B: 88.5%Mistral Large 4: 88.5%Vals AIAs of 10/26/2026
Tax Agent BenchControversy & Precedence Analysis · higher is betterTest setup not documented as identicalQwen 3.8 27B: 48.3%Mistral Large 4: 50%Vals AIAs of 10/26/2026
Code MigrationCode Quality · higher is betterTest setup not documented as identicalQwen 3.8 27B: 4.197 / 5Mistral Large 4: 4.133 / 5Vals AIAs of 10/18/2026
EMBDCF · higher is betterTest setup not documented as identicalQwen 3.8 27B: 50%Mistral Large 4: 35.9%Vals AIAs of 10/20/2026
Finance Agent (v2)Comparables · higher is betterTest setup not documented as identicalQwen 3.8 27B: 38.9%Mistral Large 4: 48.6%Vals AIAs of 10/20/2026
IOIIOI 2026 · higher is betterTest setup not documented as identicalQwen 3.8 27B: 36.8%Mistral Large 4: 47.8%Vals AIAs of 10/24/2026
Legal Research BenchBusiness & Commercial · All-pass · higher is betterTest setup not documented as identicalQwen 3.8 27B: 41.9%Mistral Large 4: 25.6%Vals AIAs of 10/26/2026
Tax Agent BenchCurrent & Temporal Analysis · higher is betterTest setup not documented as identicalQwen 3.8 27B: 43.1%Mistral Large 4: 46.6%Vals AIAs of 10/26/2026
EMBLBO · higher is betterTest setup not documented as identicalQwen 3.8 27B: 53.7%Mistral Large 4: 57.5%Vals AIAs of 10/20/2026
Finance Agent (v2)Disclosure Analysis · higher is betterTest setup not documented as identicalQwen 3.8 27B: 60.3%Mistral Large 4: 63.9%Vals AIAs of 10/20/2026
Legal Research BenchBusiness & Commercial · Weighted partial credit · higher is betterTest setup not documented as identicalQwen 3.8 27B: 86.8%Mistral Large 4: 86%Vals AIAs of 10/26/2026
Tax Agent BenchCurrent & Temporal Analysis · higher is betterTest setup not documented as identicalQwen 3.8 27B: 58.6%Mistral Large 4: 64.2%Vals AIAs of 10/26/2026
EMBM&A · higher is betterTest setup not documented as identicalQwen 3.8 27B: 64.8%Mistral Large 4: 51.9%Vals AIAs of 10/20/2026
Finance Agent (v2)Earnings Analysis · higher is betterTest setup not documented as identicalQwen 3.8 27B: 70.9%Mistral Large 4: 76.5%Vals AIAs of 10/20/2026
Legal Research BenchCivil Litigation · All-pass · higher is betterTest setup not documented as identicalQwen 3.8 27B: 25.9%Mistral Large 4: 33.3%Vals AIAs of 10/26/2026
Tax Agent BenchFact-Pattern Analysis · higher is betterTest setup not documented as identicalQwen 3.8 27B: 34.9%Mistral Large 4: 31.8%Vals AIAs of 10/26/2026
EMBMisc · higher is betterTest setup not documented as identicalQwen 3.8 27B: 59.1%Mistral Large 4: 53.9%Vals AIAs of 10/20/2026
Finance Agent (v2)Financial Modeling · higher is betterTest setup not documented as identicalQwen 3.8 27B: 19.3%Mistral Large 4: 20.5%Vals AIAs of 10/20/2026
Legal Research BenchCivil Litigation · Weighted partial credit · higher is betterTest setup not documented as identicalQwen 3.8 27B: 83.1%Mistral Large 4: 85.8%Vals AIAs of 10/26/2026
Tax Agent BenchFact-Pattern Analysis · higher is betterTest setup not documented as identicalQwen 3.8 27B: 57.7%Mistral Large 4: 57.2%Vals AIAs of 10/26/2026
EMBOperating · higher is betterTest setup not documented as identicalQwen 3.8 27B: 57%Mistral Large 4: 62.7%Vals AIAs of 10/20/2026
Finance Agent (v2)General Qualitative Analysis · higher is betterTest setup not documented as identicalQwen 3.8 27B: 62.1%Mistral Large 4: 65.4%Vals AIAs of 10/20/2026
Legal Research BenchConstitutional / Civil Rights · All-pass · higher is betterTest setup not documented as identicalQwen 3.8 27B: 26.3%Mistral Large 4: 34.2%Vals AIAs of 10/26/2026
Tax Agent BenchForms & Filings · higher is betterTest setup not documented as identicalQwen 3.8 27B: 16.7%Mistral Large 4: 26.7%Vals AIAs of 10/26/2026
EMBScratch · higher is betterTest setup not documented as identicalQwen 3.8 27B: 64.1%Mistral Large 4: 64.7%Vals AIAs of 10/20/2026
Finance Agent (v2)General Quantitative Analysis · higher is betterTest setup not documented as identicalQwen 3.8 27B: 66.9%Mistral Large 4: 75.7%Vals AIAs of 10/20/2026
Legal Research BenchConstitutional / Civil Rights · Weighted partial credit · higher is betterTest setup not documented as identicalQwen 3.8 27B: 82%Mistral Large 4: 88.1%Vals AIAs of 10/26/2026
Tax Agent BenchForms & Filings · higher is betterTest setup not documented as identicalQwen 3.8 27B: 62.4%Mistral Large 4: 71.7%Vals AIAs of 10/26/2026
EMBTemplate · higher is betterTest setup not documented as identicalQwen 3.8 27B: 54.7%Mistral Large 4: 48.1%Vals AIAs of 10/20/2026
Finance Agent (v2)Market Analysis · higher is betterTest setup not documented as identicalQwen 3.8 27B: 49.6%Mistral Large 4: 63%Vals AIAs of 10/20/2026
Legal Research BenchCriminal · All-pass · higher is betterTest setup not documented as identicalQwen 3.8 27B: 26.3%Mistral Large 4: 31.6%Vals AIAs of 10/26/2026
Tax Agent BenchNumbers & Calculations · higher is betterTest setup not documented as identicalQwen 3.8 27B: 32.3%Mistral Large 4: 40%Vals AIAs of 10/26/2026
Finance Agent (v2)Precedents · higher is betterTest setup not documented as identicalQwen 3.8 27B: 21.6%Mistral Large 4: 26.2%Vals AIAs of 10/20/2026
Legal Research BenchCriminal · Weighted partial credit · higher is betterTest setup not documented as identicalQwen 3.8 27B: 79.6%Mistral Large 4: 83.5%Vals AIAs of 10/26/2026
Tax Agent BenchNumbers & Calculations · higher is betterTest setup not documented as identicalQwen 3.8 27B: 66.2%Mistral Large 4: 78.4%Vals AIAs of 10/26/2026
Legal Research BenchFamily · All-pass · higher is betterTest setup not documented as identicalQwen 3.8 27B: 22.7%Mistral Large 4: 9.1%Vals AIAs of 10/26/2026
Tax Agent BenchRule & Source Lookup · higher is betterTest setup not documented as identicalQwen 3.8 27B: 50%Mistral Large 4: 33.3%Vals AIAs of 10/26/2026
Legal Research BenchFamily · Weighted partial credit · higher is betterTest setup not documented as identicalQwen 3.8 27B: 69.2%Mistral Large 4: 76.4%Vals AIAs of 10/26/2026
Tax Agent BenchRule & Source Lookup · higher is betterTest setup not documented as identicalQwen 3.8 27B: 68%Mistral Large 4: 60.8%Vals AIAs of 10/26/2026
Legal Research BenchHealth · All-pass · higher is betterTest setup not documented as identicalQwen 3.8 27B: 63.3%Mistral Large 4: 50%Vals AIAs of 10/26/2026
Tax Agent BenchWeighted Pass · higher is betterTest setup not documented as identicalQwen 3.8 27B: 60%Mistral Large 4: 63.3%Vals AIAs of 10/26/2026
Legal Research BenchHealth · Weighted partial credit · higher is betterTest setup not documented as identicalQwen 3.8 27B: 94.3%Mistral Large 4: 92.4%Vals AIAs of 10/26/2026
Legal Research BenchImmigration · All-pass · higher is betterTest setup not documented as identicalQwen 3.8 27B: 18.2%Mistral Large 4: 27.3%Vals AIAs of 10/26/2026
Legal Research BenchImmigration · Weighted partial credit · higher is betterTest setup not documented as identicalQwen 3.8 27B: 78%Mistral Large 4: 83.4%Vals AIAs of 10/26/2026

Compare another matchup

More matchups

Related LLM comparisons

Compare Qwen 3.8 27B and Mistral Large 4 with other leading models.

View the complete LLM comparison