How much does Qwen3 30B A3B Thinking 2507 cost?

Qwen3 30B A3B Thinking 2507 costs $0.13 per million input tokens and $1.56 per million output tokens. For a typical conversation (~2,000 tokens), that's approximately $0.002 per message.

What benchmarks has Qwen3 30B A3B Thinking 2507 been tested on?

Qwen3 30B A3B Thinking 2507 has been evaluated on 6 benchmarks. Top scores: OpenCompass — IFEval: 89.7, OpenCompass — AIME2025: 86.8, OpenCompass — MMLU-Pro: 79.5.

Is Qwen3 30B A3B Thinking 2507 open source?

Yes, Qwen3 30B A3B Thinking 2507 is open source.

How does Qwen3 30B A3B Thinking 2507 compare to GPT-4 (older v0314)?

Qwen3 30B A3B Thinking 2507 has an average score of 63.5 while GPT-4 (older v0314) scores 63.7. GPT-4 (older v0314) slightly outperforms Qwen3 30B A3B Thinking 2507 overall. Qwen3 30B A3B Thinking 2507 costs $0.13/1M input vs GPT-4 (older v0314) at $30.00/1M input. See full comparison →

Home/Models/Qwen3 30B A3B Thinking 2507

Qwen3 30B A3B Thinking 2507

Name: Qwen3 30B A3B Thinking 2507
Price: 0.13 USD
Author: Alibaba Qwen

by Alibaba Qwen · Released Aug 2025

Open Source

63.5

avg score

Rank #70

Compare

Better than 74% of all models

Context

131K tokens (~66 books)

Input $/1M

$0.13

Output $/1M

$1.56

Type

text

License

Open Source

Benchmarks

6 tested

Data updated today

About

Qwen3-30B-A3B-Thinking-2507 is a 30B parameter Mixture-of-Experts reasoning model optimized for complex tasks requiring extended multi-step thinking. The model is designed specifically for “thinking mode,” where internal reasoning traces are separated...

Tested on 6 benchmarks with 66.2% average. Top scores: OpenCompass — IFEval (89.7%), OpenCompass — AIME2025 (86.8%), OpenCompass — MMLU-Pro (79.5%).

Looking for similar performance at lower cost?
Gemma 4 31B scores 63.9 (101% as good) at $0.12/1M input · 8% cheaper

Capabilities

coding

59.4

#56 globally

math

86.8

#12 globally

knowledge

53.8

#94 globally

language

89.7

#16 globally

Benchmark Scores

Compare All

Tested on 6 benchmarks · Ranked across 4 categories

Score Distribution (all 274 models)

0255075100

▲ You are here

codingCompare coding →

OpenCompass — LiveCodeBenchV6

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

59.4—

mathCompare math →

OpenCompass — AIME2025

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

86.8—

knowledgeCompare knowledge →

OpenCompass — MMLU-Pro

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

79.5—

OpenCompass — GPQA-Diamond

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

70.1—

OpenCompass — HLE

OpenCompass evaluation of Humanitys Last Exam. Expert-level cross-discipline knowledge test.

11.7—

Quick compare:

vs GPT-4 (older v0314)

vs GPT-5.1-Codex-Mini

vs Gemma 4 31B

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

Model Family · Alibaba Qwen Qwen 3

Qwen3 14BApr 2025

$0.06/M in41Kctx

Qwen3 235B A22BApr 2025

56.4+56.4

$0.46/M in(+0.40)131Kctx(+90K)8 benchmarks

Qwen3 235B A22B Instruct 2507Jul 2025

48.5-7.9

$0.07/M in(-0.38)262Kctx(+131K)20 benchmarks

Qwen3 235B A22B Thinking 2507Jul 2025

55.9+7.4

$0.15/M in(+0.08)131Kctx(-131K)24 benchmarks

Qwen3 30B A3BApr 2025

$0.09/M in(-0.06)41Kctx(-90K)1 benchmark

Qwen3 30B A3B Instruct 2507Jul 2025

55.3+55.3

$0.09/M in262Kctx(+221K)7 benchmarks

Qwen3 30B A3B Thinking 2507Aug 2025

66.2+10.9

$0.08/M in(-0.01)131Kctx(-131K)6 benchmarks

Qwen3 32BApr 2025

58.2-8.0

$0.08/M in41Kctx(-90K)8 benchmarks

Qwen3 4B (free)Apr 2025

$0.00/M in(-0.08)41Kctx

Qwen3 8BApr 2025

56.5+56.5

$0.05/M in(+0.05)41Kctx6 benchmarks

Qwen3 Coder 30B A3B InstructJul 2025

$0.07/M in(+0.02)160Kctx(+119K)

Qwen3 Coder 480B A35BJul 2025

$0.22/M in(+0.15)262Kctx(+102K)

Qwen3 Coder 480B A35B (free)Jul 2025

$0.00/M in(-0.22)262Kctx(0K)3 benchmarks

Qwen3 Coder FlashSep 2025

$0.20/M in(+0.20)1.0Mctx(+738K)

Qwen3 Coder NextFeb 2026

$0.12/M in(-0.08)262Kctx(-738K)3 benchmarks

Qwen3 Coder PlusSep 2025

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

Qwen3 MaxSep 2025

58.3+58.3

$0.78/M in(+0.13)262Kctx(-738K)8 benchmarks

Qwen3 Max ThinkingFeb 2026

$0.78/M in262Kctx3 benchmarks

Qwen3 Next 80B A3B InstructSep 2025

54.4+54.4

$0.09/M in(-0.69)262Kctx18 benchmarks

Qwen3 Next 80B A3B Instruct (free)Sep 2025

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

Qwen3 Next 80B A3B ThinkingSep 2025

61.6+61.6

$0.10/M in(+0.10)131Kctx(-131K)20 benchmarks

Qwen3 VL 235B A22B InstructSep 2025

$0.20/M in(+0.10)262Kctx(+131K)1 benchmark

Qwen3 VL 235B A22B ThinkingSep 2025

$0.26/M in(+0.06)131Kctx(-131K)1 benchmark

Qwen3 VL 30B A3B InstructOct 2025

$0.13/M in(-0.13)131Kctx

Qwen3 VL 30B A3B ThinkingOct 2025

$0.13/M in131Kctx

Qwen3 VL 32B InstructOct 2025

$0.10/M in(-0.03)131Kctx

Qwen3 VL 8B InstructOct 2025

$0.08/M in(-0.02)131Kctx

Qwen3 VL 8B ThinkingOct 2025

$0.12/M in(+0.04)131Kctx

See the full Qwen 3 family →

Similar Models

Links

Info

Alibaba Qwen Pricing explorer Developers · API

Research

Documentation

Community

Source Code

BenchGecko API

qwen3-30b-a3b-thinking-2507

Specifications

Typetext
Context131K tokens (~66 books)
ReleasedAug 2025
LicenseOpen Source
StatusActive
Cost / Message~$0.002

Available On

Alibaba Qwen$0.13

Frequently Asked Questions

Qwen3 30B A3B Thinking 2507 is an open-source text AI model by Alibaba Qwen, released in August 2025. It has an average benchmark score of 63.5. Context window: 131K tokens.

Benchmarks

OpenCompass — IFEval OpenCompass — AIME2025 OpenCompass — MMLU-Pro OpenCompass — GPQA-Diamond OpenCompass — LiveCodeBenchV6

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

Qwen3 30B A3B Thinking 2507

Frequently Asked Questions

Related Models

Benchmarks

Related Pages