How much does Qwen3 8B cost?

Qwen3 8B costs $0.12 per million input tokens and $0.46 per million output tokens. For a typical conversation (~2,000 tokens), that's approximately $0.001 per message.

What benchmarks has Qwen3 8B been tested on?

Qwen3 8B has been evaluated on 12 benchmarks. Top scores: OpenCompass — IFEval: 85.6, OpenCompass — MMLU-Pro: 72.1, OpenCompass — AIME2025: 66.2.

How does Qwen3 8B compare to Llama 3.2 90B?

Qwen3 8B has an average score of 36.3 while Llama 3.2 90B scores 36.3. Llama 3.2 90B slightly outperforms Qwen3 8B overall. See full comparison →

Home/Models/Qwen3 8B

Qwen3 8B

Name: Qwen3 8B · benchmarks, pricing and providers
Creator: BenchGecko

by Alibaba Qwen · Released Apr 2025

Open Source

36.3

avg score

Rank #217

Compare

Better than 30% of all models

Context

131K tokens (~66 books)

Input $/1M

$0.12

Output $/1M

$0.46

Type

text

License

Open Source

Benchmarks

12 tested

Data as of 2026-10-05

About

Qwen3-8B is a dense 8.2B parameter causal language model from the Qwen3 series, designed for both reasoning-heavy tasks and efficient dialogue. It supports seamless switching between "thinking" mode for math,...

Tested on 12 benchmarks · BenchGecko score 36.3. Top scores: OpenCompass — IFEval (85.6%), OpenCompass — MMLU-Pro (72.1%), OpenCompass — AIME2025 (66.2%).

Looking for similar performance at lower cost?
Qwen3.5-Flash scores 36.5 (101% as good) at $0.07/1M input · 44% cheaper

Capabilities

coding

50.1

#91 globally

math

61.1

#80 globally

knowledge

40.3

#202 globally

general

21.6

#159 globally

language

85.6

#34 globally

Benchmark Scores

Compare All

Tested on 12 benchmarks · Ranked across 5 categories

Score Distribution (all 312 models)

0255075100

▲ You are here

codingCompare coding →

OpenCompass — LiveCodeBenchV6

OpenCompass Live Code Bench v6. Fresh competitive programming problems to evaluate code generation without memorization.

50.1·

mathCompare math →

OpenCompass — AIME2025

OpenCompass evaluation on AIME 2025 problems. Tests mathematical reasoning on fresh competition problems.

66.2·

OTIS Mock AIME 2024-2025

Mock AIME (American Invitational Mathematics Exam) problems from OTIS. Tests mathematical competition performance.

56.1·

knowledgeCompare knowledge →

OpenCompass — MMLU-Pro

OpenCompass MMLU-Pro evaluation. Harder knowledge test with more answer choices.

72.1·

Fiction.LiveBench

LiveBench fiction analysis. Tests literary comprehension and creative text understanding.

62.1·

OpenCompass — GPQA-Diamond

OpenCompass evaluation of GPQA Diamond. PhD-level science questions from the hardest subset.

59.7·

Quick compare:

vs Llama 3.2 90B

vs Gemini 1.5 Pro (May 2024)

vs Qwen2.5-Max

Excellent (85+) Good (70-85) Average (50-70) Below (<50)

Model Family · Alibaba Qwen Qwen 3

Qwen3 14BApr 2025

39.2

$0.12/M in131Kctx6 benchmarks

Qwen3 235B A22BApr 2025

55.2+16.0

$0.46/M in(+0.34)131Kctx10 benchmarks

Qwen3 235B A22B Instruct 2507Jul 2025

44.9-10.3

$0.09/M in(-0.36)262Kctx(+131K)22 benchmarks

Qwen3 235B A22B Thinking 2507Jul 2025

55.3+10.4

$0.23/M in(+0.14)131Kctx(-131K)27 benchmarks

Qwen3 30B A3BApr 2025

34.0-21.3

$0.12/M in(-0.11)131Kctx9 benchmarks

Qwen3 30B A3B Instruct 2507Jul 2025

40.1+6.1

$0.05/M in(-0.07)262Kctx(+131K)7 benchmarks

Qwen3 30B A3B Thinking 2507Aug 2025

63.5+23.4

$0.20/M in(+0.15)82Kctx(-180K)6 benchmarks

Qwen3 32BApr 2025

47.8-15.7

$0.08/M in(-0.12)131Kctx(+49K)14 benchmarks

Qwen3 4B (free)Apr 2025

$0.00/M in(-0.08)41Kctx(-90K)

Qwen3 8BApr 2025

36.3-11.5

$0.12/M in(+0.12)131Kctx(+90K)12 benchmarks

Qwen3 Coder 30B A3B InstructJul 2025

$0.07/M in(-0.05)262Kctx(+131K)

Qwen3 Coder 480B A35BJul 2025

$0.30/M in(+0.23)262Kctx

Qwen3 Coder 480B A35B (free)Jul 2025

$0.00/M in(-0.30)1.0Mctx(+786K)3 benchmarks

Qwen3 Coder FlashSep 2025

$0.20/M in(+0.20)1.0Mctx(-49K)

Qwen3 Coder NextFeb 2026

$0.12/M in(-0.08)262Kctx(-738K)12 benchmarks

Qwen3 Coder PlusSep 2025

$0.65/M in(+0.53)1.0Mctx(+738K)

Qwen3 MaxSep 2025

49.2+12.9

$0.78/M in(+0.13)262Kctx(-738K)13 benchmarks

Qwen3 Max ThinkingFeb 2026

$0.78/M in262Kctx3 benchmarks

Qwen3 Next 80B A3B InstructSep 2025

49.2

$0.09/M in(-0.69)262Kctx26 benchmarks

Qwen3 Next 80B A3B Instruct (free)Sep 2025

$0.00/M in(-0.09)262Kctx3 benchmarks

Qwen3 Next 80B A3B ThinkingSep 2025

56.8+7.6

$0.15/M in(+0.15)262Kctx20 benchmarks

Qwen3 VL 235B A22B InstructSep 2025

$0.21/M in(+0.06)262Kctx1 benchmark

Qwen3 VL 235B A22B ThinkingSep 2025

$0.40/M in(+0.19)131Kctx(-131K)1 benchmark

Qwen3 VL 30B A3B InstructOct 2025

$0.15/M in(-0.25)262Kctx(+131K)

Qwen3 VL 30B A3B ThinkingOct 2025

$0.20/M in(+0.05)262Kctx

Qwen3 VL 32B InstructOct 2025

$0.10/M in(-0.10)131Kctx(-131K)

Qwen3 VL 8B InstructOct 2025

$0.12/M in(+0.01)262Kctx(+131K)

Qwen3 VL 8B ThinkingOct 2025

$0.18/M in(+0.06)131Kctx(-131K)

See the full Qwen 3 family →

Recently Happened

Qwen3 8B pricing increased 134%

Jul 2, 2026

Similar Models

Llama 3.2 90B

Frequently Asked Questions

Qwen3 8B is an open-source text AI model by Alibaba Qwen, released in April 2025. It has an average benchmark score of 36.3. Context window: 131K tokens.

Benchmarks

OpenCompass — IFEval OpenCompass — MMLU-Pro OpenCompass — AIME2025 Fiction.LiveBench OpenCompass — GPQA-Diamond

Alibaba Qwen · Provider Alibaba Qwen · Economy All Models Compare Models Pricing Developers · API Context Window · Glossary

Key facts · as of 2026-10-05

Qwen3 8B by Alibaba Qwen. BenchGecko score 36.3, rank 215 of 312 scored models (normalized average of public benchmark scores).
List price $0.12 input · $0.46 output per 1M tokens (as of 2026-10-05).
Sold by 1 provider (as of 2026-10-05): Alibaba $0.12 in / $0.46 out.

How to cite · data as of 2026-10-05

Qwen3 8B · benchmarks, pricing and providers. BenchGecko, data as of 2026-10-05. https://benchgecko.ai/model/qwen3-8b

Credit "Source: BenchGecko" with a link. Prices per provider and Gecko Tests are BenchGecko data (CC BY 4.0); benchmark scores keep their original source, listed in the JSON. JSON · llms.txt · MCP

Qwen3 8B

Frequently Asked Questions

Related Models

Benchmarks

Related Pages

Key facts · as of 2026-10-05

How to cite · data as of 2026-10-05