Compare / Qwen3-4B-Instruct-2507 vs GPT-5.2
Qwen3-4B-Instruct-2507 vs GPT-5.2
Which is cheaper, which is more capable, and which does better at each kind of work, from 18 results on 3 sources that measured both models.
Of the 18 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.
Results as of 9 October 2026 (catalogue release 2026-10-09-dd620e778ac5). See also Alibaba vs OpenAI, every model from both labs.
Qwen3-4B-Instruct-2507 or GPT-5.2? The short answer
Pick GPT-5.2: it scores higher (Intelligence Index 30.4 against 6.7) at about the same price.
- GPT-5.2 scores higher on the Artificial Analysis Intelligence Index: 30.4 against 6.7.
- GPT-5.2 is newer: released 11 Dec 2025, against 6 Aug 2025 for Qwen3-4B-Instruct-2507.
- On third-party benchmarks, on the results both models have: GPT-5.2 does better for agents and tool use, reasoning and knowledge and writing.
Head to head
Every fact side by side
| Qwen3-4B-Instruct-2507 | GPT-5.2 | |
|---|---|---|
| Price per million tokens | Not on sale | $1.75 in · $14.00 out |
| Blended (3 in : 1 out) | – | $4.81 |
| Cheapest host, blended | – | $4.81 (Azure) |
| Intelligence Index | 6.7 | 30.4 |
| Spring Prompt overall | – | – |
| Output speed, tokens a second | – | – |
| Context window | – | 400,000 tokens |
| Released | 6 Aug 2025 | 11 Dec 2025 (newer) |
| Weights | Closed | Closed |
| Developer based in | China | the US |
Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.
Which is better for what
Each model's rank among the models on sale today, for each kind of work, from our own benchmarks and licensed sources. Further right is better; the better rank is in green.
From our research
Every shared result
Each model's best result on every measure both have, grouped by benchmark, with the setting that produced it. Under each measure, every model that benchmark has measured is a grey tick, best to the right, so you can see where the two sit in the field. The headline measures of each benchmark come first; the rest are under "Show all". Green marks the better value; where a source reports ranges that overlap, neither is marked.
| Measure | Qwen3-4B-Instruct-2507 | GPT-5.2 |
|---|---|---|
| Artificial Analysis reported by Artificial Analysis · 3 results | ||
| Artificial Analysis Intelligence Indexindex score, higher is better 222 · 49 of 243 measured | 6.7 | 30.4xhigh reasoning |
| GPQA Diamond% of questions, higher is better 198 · 40 of 231 measured | 51.7% | 90.3%xhigh reasoning |
| Humanity's Last Exam% of questions, higher is better 202 · 50 of 242 measured | 4.5% | 37.7%xhigh reasoning |
| Berkeley Function Calling Leaderboard (BFCL) V4 reported by Berkeley Function Calling Leaderboard (BFCL) V4 · 3 results | ||
| Overall accuracy · BFCL% correct, higher is better 40 · 14 of 81 measured | 35.7%bfcl fc | 55.9%bfcl fc |
| Irrelevance detection · BFCL% correct, higher is better 30 · 19 of 81 measured | 84.9%bfcl fc | 87.3%bfcl prompt |
| Memory · BFCL% correct, higher is better 26 · 10 of 81 measured | 23.9%bfcl prompt | 45.8%bfcl fc |
| UGI Leaderboard reported by UGI Leaderboard · 3 results | ||
| Writing score · UGIscore out of 100, higher is better 154 · 78 of 230 measured | 29.9 | 49.4high reasoning |
| Requested-length error · UGI% off the requested word count, lower is better 25 · 214 of 230 measured | 8.0% | 76.0%high reasoning |
| Style adherence · UGIscore from 0 to 1, higher is better 117 · 17 of 230 measured | 0.35 | 0.40low reasoning |
Show all 18 results (9 more)
| Measure | Qwen3-4B-Instruct-2507 | GPT-5.2 |
|---|---|---|
| Artificial Analysis reported by Artificial Analysis · 4 results | ||
| IFBench% of instructions, higher is better 164 · 26 of 195 measured | 33.5% | 75.4%xhigh reasoning |
| Long-context reasoning (AA-LCR)% of questions, higher is better 214 · 19 of 233 measured | 11.3% | 82.7%xhigh reasoning |
| Terminal-Bench Hard% of tasks, higher is better 156 · 15 of 190 measured | 4.5% | 47.0%xhigh reasoning |
| τ²-bench telecom% of tasks, higher is better 146 · 60 of 193 measured | 26.6% | 84.8%xhigh reasoning |
| Berkeley Function Calling Leaderboard (BFCL) V4 reported by Berkeley Function Calling Leaderboard (BFCL) V4 · 5 results | ||
| Multi-turn tasks · BFCL% correct, higher is better 40 · 22 of 81 measured | 22.1%bfcl fc | 43.8%bfcl prompt |
| Relevance detection · BFCL% correct, higher is better 26 · 49 of 81 measured | 87.5%bfcl fc | 75.0%bfcl fc |
| Single-turn calls (curated) · BFCL% correct, higher is better 21 · 53 of 81 measured | 87.9%bfcl fc | 81.9%bfcl fc |
| Single-turn calls (user-contributed) · BFCL% correct, higher is better 28 · 47 of 81 measured | 76.4%bfcl fc | 70.4%bfcl fc |
| Web search · BFCL% correct, higher is better 44 · 10 of 81 measured | 4.5%bfcl prompt | 75.5%bfcl fc |
Compare them with other models
Quick answers
Is Qwen3-4B-Instruct-2507 better than GPT-5.2?
Pick GPT-5.2: it scores higher (Intelligence Index 30.4 against 6.7) at about the same price.
Which is better for agents and tool use, Qwen3-4B-Instruct-2507 or GPT-5.2?
GPT-5.2 ranks 57 of 165 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 97 for Qwen3-4B-Instruct-2507.
Which is better for reasoning and knowledge, Qwen3-4B-Instruct-2507 or GPT-5.2?
GPT-5.2 ranks 48 of 199 models on sale for hard questions across science, maths and general knowledge, against 172 for Qwen3-4B-Instruct-2507.
Which is better for writing, Qwen3-4B-Instruct-2507 or GPT-5.2?
GPT-5.2 ranks 78 of 189 models on sale for what people prefer in blind comparisons, and judged writing quality, against 149 for Qwen3-4B-Instruct-2507.
Run Qwen3-4B-Instruct-2507 and GPT-5.2 on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.