Qwen3.8 27B vs Gemma 4 26B A4B
Qwen3.8 27B scores higher on 3 of the 3 benchmarks both models publish. The widest gap is on Humanity's Last Exam, where Qwen3.8 27B leads by 22.1 points.
Qwen3.8 27BAlibaba Qwen · OpenLLM Buddy catalog83.2composite
Gemma 4 26B A4BGoogle · OpenLLM Buddy catalog72.9compositeShared benchmarks
Raw published scores; the higher score on each row is highlighted. LMArena Elo is a human-preference vote and is not counted as a win.
- GPQA Diamond89.282.3
- Humanity's Last Exam30.88.7
no tools
- LiveCodeBench90.377.1
LiveCodeBench v6LiveCodeBench v6
- LMArena Text Elo · not counted14361438
LMArena text overall, 'Qwen3.8-27B', ±5, as of 2026-10-27LMArena text overall, 'Gemma 4 26B A4B', ±7, as of 2026-10-27
Should you pick Qwen3.8 27B or Gemma 4 26B A4B?
Pick Qwen3.8 27B if…
- You care most about GPQA Diamond, Humanity's Last Exam, LiveCodeBench, where it scores at least 2 points higher.
- You want it running in minutes on a dedicated GPU through OpenLLM Buddy, in an uncensored build.
Pick Gemma 4 26B A4B if…
- You need fast responses: about 113 tokens per second through APIs, against 41.
- Cost matters: $0.600 per million output tokens against $2.55.
- You want it running in minutes on a dedicated GPU through OpenLLM Buddy.
Hardware and cost to run
Qwen3.8 27B
At 4-bit it needs about 19 GB, which fits a single RTX 4090, so it can run on a home workstation with llama.cpp, Ollama, or vLLM. Full BF16 precision needs about 62 GB (1× H100 (80 GB)).
On OpenLLM Buddy: Qwen3.8-27B-Uncensored-FP8 runs on H100 80GB as an FP8 build, with an OpenAI-compatible API and no per-token fees. Prices start at $53 USD for 11 hours.
Deploy Qwen3.8 27BGemma 4 26B A4B
At 4-bit it needs about 17 GB, which fits a single RTX 4090, so it can run on a home workstation with llama.cpp, Ollama, or vLLM. Full BF16 precision needs about 58 GB (1× H100 (80 GB)).
On OpenLLM Buddy: Gemma 4 26B A4B, with an OpenAI-compatible API and no per-token fees. Prices start at $13 USD for 11 Hours.
Deploy Gemma 4 26B A4BCategory profile
Category index uses every test each model reports, not only the shared ones.
Specs side by side
| Qwen3.8 27B | Gemma 4 26B A4B | |
|---|---|---|
| Lab | Alibaba Qwen | |
| Weights | Open | Open |
| Released | 2026-08-14 | 2026-04-02 |
| Parameters | 27B dense | 25.2B total, 3.8B active (MoE) |
| Context | 262K (up to 1M) | 256K |
| License | Apache-2.0 | Apache-2.0 |
| 4-bit memory | 19 GB · RTX 4090 (24 GB) | 17 GB · RTX 4090 (24 GB) |
| API price | $0.425 in / $2.55 out per 1M tokens | $0.150 in / $0.600 out per 1M tokens |
| API speed | 41 tok/s | 113 tok/s |
| One-GPU speed (est.) | ≈ 52–69 tok/s on H100 80GB (FP8) | ≈ 133–160 tok/s on RTX 4090 (4-bit) |
Related comparisons
Sources
| Model | Test | Score | Setting | Source |
|---|---|---|---|---|
| Qwen3.8 27B | GPQA Diamond | 89.2 | — | Qwen3.8-27B model card (Hugging Face) |
| Gemma 4 26B A4B | GPQA Diamond | 82.3 | — | Gemma 4 31B IT model card (Hugging Face), 26B A4B column |
| Qwen3.8 27B | Humanity's Last Exam | 30.8 | — | Qwen3.8-27B model card (Hugging Face) |
| Gemma 4 26B A4B | Humanity's Last Exam | 8.7 | no tools | Gemma 4 31B IT model card (Hugging Face), 26B A4B column |
| Qwen3.8 27B | LiveCodeBench | 90.3 | LiveCodeBench v6 | Qwen3.8-27B model card (Hugging Face) |
| Gemma 4 26B A4B | LiveCodeBench | 77.1 | LiveCodeBench v6 | Gemma 4 31B IT model card (Hugging Face), 26B A4B column |
| Qwen3.8 27B | LMArena Text Elo | 1436 | LMArena text overall, 'Qwen3.8-27B', ±5, as of 2026-10-27 | Arena (LMArena) text leaderboard |
| Gemma 4 26B A4B | LMArena Text Elo | 1438 | LMArena text overall, 'Gemma 4 26B A4B', ±7, as of 2026-10-27 | Arena (LMArena) text leaderboard |
Scores are published results from model cards, lab reports, and public leaderboards, each linked to its source. Settings such as thinking mode or test version differ between sources and are shown next to each score. OpenLLM Buddy catalog models are uncensored fine-tunes; their scores are those of the official base model. Data collected 2026-10-30. How the index works