Qwen3 4B 2507 Thinking
Qwen3 4B 2507 Thinking is behind the leaders in instruction following. Too few results yet to rate reasoning, coding, agentic tasks, safety, long context, math, multimodal tasks, multilingual tasks, or factuality.
Capability profile
Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.
Too few results yet to rate Reasoning, Coding, Agentic, Safety, Long Context, Math, Multimodal, Multilingual or Factuality.
Price
$0.01input$0.03outputper million tokens
From nscale · All prices
Evidence
14results on13benchmarks
- 3 independently verified
- 11 aggregator
From 3 sources · latest Oct 7, 2026 · How verification works
Qwen3 4B 2507 Thinking benchmark results
14 results on 13 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.
39.3% behind the leader1 of 3 ranked benchmarks measured
- IFBench49.80Oct 7, 2026aa_ifbench
0 of 6 ranked benchmarks measured
- 0.00Oct 7, 2026
- GPQA Diamond66.67Oct 7, 2026gpqa
- Humanity's Last Exam6.16Oct 7, 2026aa_hle
0 of 10 ranked benchmarks measured
- Terminal-Bench Hard1.52Oct 7, 2026aa_terminalbench_hard
0 of 3 ranked benchmarks measured
- 37.33Oct 7, 2026
0 of 5 ranked benchmarks measured
- 53.03Sep 2, 2026
- 82.50Sep 2, 2026
- 53.03May 10, 2026
0 of 4 ranked benchmarks measured
- AA-Omniscience · Accuracy12.53Oct 7, 2026omniscienceAccuracy
- AA-Omniscience · Non-hallucination16.77Oct 7, 2026omniscienceNonHallucination
More results
Benchmarks outside the capability baskets. They are not ranked against a leader.
- τ²-Bench Telecom (AA run)25.44Oct 7, 2026aa_tau2
- -60.27Oct 7, 2026
- AA Intelligence8.80Oct 7, 2026aa_intelligence_index
Qwen3 4B 2507 Thinking: common questions
Who makes Qwen3 4B 2507 Thinking?
Qwen3 4B 2507 Thinking is made by Alibaba.
When was Qwen3 4B 2507 Thinking released?
Qwen3 4B 2507 Thinking was released on Aug 6, 2025, according to Artificial Analysis.
What is Qwen3 4B 2507 Thinking good at?
Qwen3 4B 2507 Thinking is behind the leaders in instruction following. Too few results yet to rate reasoning, coding, agentic tasks, safety, long context, math, multimodal tasks, multilingual tasks, or factuality.
How much does Qwen3 4B 2507 Thinking cost?
Qwen3 4B 2507 Thinking costs $0.01 per million input tokens and $0.03 per million output tokens, according to nscale. At a mix of three input tokens to one output token, it is cheaper than 99% of the 328 priced models we track.
How many benchmarks has Qwen3 4B 2507 Thinking been tested on?
We track 14 results for Qwen3 4B 2507 Thinking on 13 benchmarks from 3 sources, 3 of them independently verified. The latest was recorded on Oct 7, 2026.
About this record
Where Qwen3 4B 2507 Thinking's numbers come from, and every name it appears under.
- Tracked since
- Sep 23, 2026
- Newest source mention
- Sep 23, 2026
Where the results come from
Verification: 14 scores · 3 independently verified · 11 aggregator-attributed. How these tiers are assigned
From 3 sources on 2 sites. Artificial Analysis supplies 11 of them; the 3 independently verified results come from 1 site. Bars are coloured by trust tier.
- artificialanalysis.ai11
- matharena.ai3