Model comparison
Grok 4.20 Multi-Agent vs Qwen3.5 27B
Grok 4.20 Multi-Agent is the stronger model overall, scoring 45.5 to 41.4 on the Noometry Index. Qwen3.5 27B costs 1.9× less per token, which makes it the better buy when Grok 4.20 Multi-Agent's lead doesn't matter for your workload.
Last verified . 19 shared benchmarks.
Alibaba (Qwen)Summary
- They share 19 benchmarks with published results for both. Grok 4.20 Multi-Agent scores higher in 8 categories and Qwen3.5 27B in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4.20 Multi-Agent leads 42.4 to 26.6.
- The biggest single-benchmark swing is NYT Connections (extended): 89.6% for Grok 4.20 Multi-Agent and 47.9% for Qwen3.5 27B.
- Qwen3.5 27B is cheaper at $0.30 / $2.40 per million input/output tokens, against $1.25 / $2.50 for Grok 4.20 Multi-Agent.
- Grok 4.20 Multi-Agent accepts more context: 1M tokens versus 262K.
- Qwen3.5 27B has downloadable open weights; the other is API-only.
Side by side
| Grok 4.20 Multi-Agent | Qwen3.5 27B | |
|---|---|---|
| Provider | xAI | Alibaba (Qwen) |
| Noometry Index | 45.5 | 41.4 |
| Released | 2026-03-09 | 2026-02-23 |
| Weights | Proprietary | Open |
| Context window | 1M | 262K |
| Max output | 30K | 66K |
| Input $ / M tokens | $1.25 | $0.30 |
| Output $ / M tokens | $2.50 | $2.40 |
| Results tracked | 20 | 28 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.20 Multi-Agent leads
Grok 4.20 Multi-Agent: 42.4 (#91), Qwen3.5 27B: 38.6 (#159)
| Benchmark | Grok 4.20 Multi-Agent | Qwen3.5 27B |
|---|---|---|
| LMArena Coding | 1458 | 1427 |
| LMArena WebDev | — | 1357 |
| WeirdML | — | 39.5% |
| ALE-Bench | — | 349.45 |
Agentic & Tool Use Not comparable
Grok 4.20 Multi-Agent: —, Qwen3.5 27B: —
| Benchmark | Grok 4.20 Multi-Agent | Qwen3.5 27B |
|---|---|---|
| LMArena Search | 1204 | — |
| Vending-Bench 2 | — | 201.98 |
Reasoning Grok 4.20 Multi-Agent leads
Grok 4.20 Multi-Agent: 42.4 (#44), Qwen3.5 27B: 26.6 (#116)
| Benchmark | Grok 4.20 Multi-Agent | Qwen3.5 27B |
|---|---|---|
| NYT Connections (extended) | 89.6% | 47.9% |
| LMArena Hard Prompts | 1448 | 1414 |
| Thematic Generalization | — | 45.5% |
| DTBench | — | 82.4% |
| LMCA | — | 34% |
Math Too close to call
Grok 4.20 Multi-Agent: 36.6 (#106), Qwen3.5 27B: 36.8 (#103)
| Benchmark | Grok 4.20 Multi-Agent | Qwen3.5 27B |
|---|---|---|
| LMArena Math | 1441 | 1428 |
| MathArena Final-Answer Competitions | — | 56.7% |
Knowledge Grok 4.20 Multi-Agent leads
Grok 4.20 Multi-Agent: 39.7 (#116), Qwen3.5 27B: 37.4 (#149)
| Benchmark | Grok 4.20 Multi-Agent | Qwen3.5 27B |
|---|---|---|
| LMArena Expert | 1445 | 1428 |
| Vectara Hallucination Rate | — | 12.1% |
Multimodal Grok 4.20 Multi-Agent leads
Grok 4.20 Multi-Agent: 39.8 (#44), Qwen3.5 27B: 38.7 (#56)
| Benchmark | Grok 4.20 Multi-Agent | Qwen3.5 27B |
|---|---|---|
| LMArena Vision | 1259 | 1240 |
Multilingual Grok 4.20 Multi-Agent leads
Grok 4.20 Multi-Agent: 54.6 (#40), Qwen3.5 27B: 51.0 (#112)
| Benchmark | Grok 4.20 Multi-Agent | Qwen3.5 27B |
|---|---|---|
| LMArena Non-English | 1440 | 1390 |
| LMArena Chinese | 1475 | 1478 |
| LMArena French | 1467 | 1409 |
| LMArena German | 1456 | 1393 |
| LMArena Japanese | 1405 | 1345 |
| LMArena Korean | 1416 | 1359 |
| LMArena Russian | 1457 | 1389 |
| LMArena Spanish | 1447 | 1408 |
Instruction Following Grok 4.20 Multi-Agent leads
Grok 4.20 Multi-Agent: 75.0 (#80), Qwen3.5 27B: 73.7 (#114)
| Benchmark | Grok 4.20 Multi-Agent | Qwen3.5 27B |
|---|---|---|
| LMArena Instruction Following | 1420 | 1393 |
Long Context Too close to call
Grok 4.20 Multi-Agent: 43.7 (#84), Qwen3.5 27B: 43.1 (#103)
| Benchmark | Grok 4.20 Multi-Agent | Qwen3.5 27B |
|---|---|---|
| LMArena Longer Query | 1430 | 1413 |
Writing & Preference Grok 4.20 Multi-Agent leads
Grok 4.20 Multi-Agent: 64.1 (#57), Qwen3.5 27B: 59.5 (#106)
| Benchmark | Grok 4.20 Multi-Agent | Qwen3.5 27B |
|---|---|---|
| LMArena Text | 1450 | 1409 |
| LMArena Creative Writing | 1435 | 1362 |
| LMArena Multi-Turn | 1451 | 1410 |
Frequently asked questions
Is Grok 4.20 Multi-Agent better than Qwen3.5 27B?
Grok 4.20 Multi-Agent is the stronger model overall, scoring 45.5 to 41.4 on the Noometry Index. Qwen3.5 27B costs 1.9× less per token, which makes it the better buy when Grok 4.20 Multi-Agent's lead doesn't matter for your workload.
Which is cheaper, Grok 4.20 Multi-Agent or Qwen3.5 27B?
Qwen3.5 27B is cheaper. It lists at $0.30 per million input tokens and $2.40 per million output tokens; Grok 4.20 Multi-Agent lists at $1.25 and $2.50.
Is Grok 4.20 Multi-Agent or Qwen3.5 27B better for coding?
Grok 4.20 Multi-Agent scores higher on coding benchmarks: 42.4 versus 38.6 in the Noometry coding category.
Which has the bigger context window?
Grok 4.20 Multi-Agent does, with 1M tokens against 262K.
How many benchmarks do Grok 4.20 Multi-Agent and Qwen3.5 27B share?
19 benchmarks have published results for both models. Grok 4.20 Multi-Agent has 20 scored results on Noometry and Qwen3.5 27B has 28.