Model comparison
INTELLECT-1 vs Qwen3-4B
Qwen3-4B has enough public results to be ranked (#264); INTELLECT-1 does not yet, so treat this comparison as directional.
Last verified . 0 shared benchmarks.
Hugging Face
Alibaba (Qwen)Summary
- Qwen3-4B has downloadable open weights; the other is API-only.
Side by side
| INTELLECT-1 | Qwen3-4B | |
|---|---|---|
| Provider | Hugging Face | Alibaba (Qwen) |
| Noometry Index | — | 31.9 |
| Released | 2024-11-29 | 2025-04-29 |
| Weights | Proprietary | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 7 | 6 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Agentic & Tool Use Not comparable
INTELLECT-1: —, Qwen3-4B: 27.6 (#100)
| Benchmark | INTELLECT-1 | Qwen3-4B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 35.7% |
Reasoning Not comparable
INTELLECT-1: —, Qwen3-4B: 19.2 (#268)
| Benchmark | INTELLECT-1 | Qwen3-4B |
|---|---|---|
| Chess Puzzles | — | 4% |
| BIG-Bench Hard | 34.9% | — |
| Epoch Capabilities Index | 100.86 | — |
| HellaSwag | 71.4% | — |
| WinoGrande | 65.8% | — |
Math Not comparable
INTELLECT-1: —, Qwen3-4B: 29.7 (#240)
| Benchmark | INTELLECT-1 | Qwen3-4B |
|---|---|---|
| MathArena Final-Answer Competitions | — | 38.5% |
| OTIS Mock AIME 2024-2025 | — | 52.2% |
| GSM8K | 38.6% | — |
Knowledge Not comparable
INTELLECT-1: —, Qwen3-4B: 33.0 (#208)
| Benchmark | INTELLECT-1 | Qwen3-4B |
|---|---|---|
| GPQA Diamond | — | 52.3% |
| Vectara Hallucination Rate | — | 5.7% |
| ARC (AI2) Challenge | 54.5% | — |
| MMLU | 49.9% | — |
Frequently asked questions
Is INTELLECT-1 better than Qwen3-4B?
Qwen3-4B has enough public results to be ranked (#264); INTELLECT-1 does not yet, so treat this comparison as directional.
How many benchmarks do INTELLECT-1 and Qwen3-4B share?
0 benchmarks have published results for both models. INTELLECT-1 has 7 scored results on Noometry and Qwen3-4B has 6.